├── README.md └── test_predict.ipynb /README.md: -------------------------------------------------------------------------------- 1 | 2 | # Simultaneous Vehicle Detection and Classification Model based on Deep YOLO Networks 3 | Detect and classify iranian vehicle make and model based on deep cnn neural networks with keras framework. 4 | Implementation and review of the article [IEEE](https://ieeexplore.ieee.org/abstract/document/9116922) 5 | 6 | ## Abstract 7 | 8 | Due to the rapid growth of vehicles, traffic monitoring and tracking systems in the last decade, vehicle detection and extracting information such as vehicle type and model and car plate recognition are the important issues of the day and many efforts have been made using different methods to identify vehicles. Given the high number of vehicles and similarities of classes, it is a difficult task to find an accurate and rapid approach to differentiate available classes and classify them. In this article we propose a new approach which simultaneously detects vehicles and identifies their type. Two models are trained in this paper. The first model is based on CNN networks to extract features and detect vehicle models, and the second model which is the main contribution of this article is based on YOLO (You Only Look Once) algorithm and SSD to detect vehicle location in the image. To train and test this approach, we collect 150,000 images of 115 domestic and foreign vehicle classes from Iranian websites. Hence, the images have large variations in image size, illumination and pose. The experimental results on the accrued dataset shows that the proposed method is able to correctly classify 91% of the vehicles in uncontrolled conditions. 9 | 10 | ## PROPOSED METHOD 11 | There are various methods to extract and classify the features available in a vehicle image. In this study, two different scenarios were employed to classify and identify vehicle type and its location in the image. The first scenario involves combining an SSD object detection network in the image to determine the vehicle location and a ResNet convolutional class network to identify vehicle type. In the second scenario, which is the approach proposed in this paper, the YOLO algorithm is used to train an end-to-end network for the seamless and direct detection of all vehicles in the image. 12 | 13 | 14 | 15 | 16 | ## Data Collection and dataset creation 17 | To collect data and create this dataset of common vehicles in Iran, a script was implemented which surveyed car dealing websites and extracted and saved images of vehicles in available categories. Then, by examining the number of images of cars in each class, categories with over 400 images were selected for testing. A total of 95 vehicle classes were prepared to conduct the experiment. Images removed from other classes are included in class 115, which has been considered for making models other than the models in the main classes. 18 | 19 | | Row | Vehicle | Model | Number | 20 | |--|--|--|--| 21 | |1 | Saipa | Pride | 14833 | 22 | |2 | IranKhodro| Peykan| 7442| 23 | |3 | Peugeot| Pride | 7367| 24 | |... | ...| ... | ...| 25 | 26 | Ultimately, a dataset containing 148414 images of various classes was prepared. Table I displays 3 vehicles in the database by brand, model and number. 27 | 28 | 29 | 30 | 31 | To download the dataset, please send your request to my email amirmg.1375@gmail.com. 32 | 33 | ## RESULTS 34 | In experiment 1, a 91.0% accuracy was obtained with the dataset introduced at the beginning of this section including 96 classes of common vehicles in Iran. A 91.2% 35 | accuracy was obtained in experiment 2. As we observe, the second method is more accurate than the first one which represents that pre-processing and cutting images had a positive effect on accuracy improvement. Table II show some of the true and false results obtained from experiments 1 and 2. The accuracy value reported in the second scenario in Table II is based on the value of the detected confidence for each class and applying the specified threshold on the values. 36 | 37 | | | Scenario 1 | Scenario 1 | Scenario 2 | 38 | |----------------|---------------|---------------|-------------| 39 | | | Experiment 1 | Experiment 2 |Experiment 1 | 40 | | Accuracy | 89 | 90.5 | 75.03 | 41 | | Mean time (second) | 1.378 | 1.378 | 0.08 | 42 | | Input image size | 324×324 | 424×424 | 416×416 | 43 | | mAP | - | - | 0.73 | 44 | | IoU threshold | - | - | 0.66 | 45 | | Precision | 0.9 | 0.91 | 0.72 | 46 | | Recall | 0.9 | 0.88 | 0.73 | 47 | | F1-score | 0.9 | 0.89 | 0.72 | 48 | 49 | 50 | 51 | 52 | 53 | 54 | 55 | 56 | ## How to run the model 57 | 58 | First download the pretrained cnn model from this link address 59 | 60 | https://drive.google.com/open?id=13x7I9UyVLSReSJAmoroRNjqEUyyYxNvz 61 | 62 | 63 | Then run the `test_predict.py` notebooke and get the results. 64 | 65 | ## Citation 66 | ``` 67 | @INPROCEEDINGS{9116922, 68 | author={Ghoreyshi, Amir Mohammad and AkhavanPour, Alireza and Bossaghzadeh, Alireza}, 69 | booktitle={2020 International Conference on Machine Vision and Image Processing (MVIP)}, 70 | title={Simultaneous Vehicle Detection and Classification Model based on Deep YOLO Networks}, 71 | year={2020}, 72 | volume={}, 73 | number={}, 74 | pages={1-6}, 75 | doi={10.1109/MVIP49855.2020.9116922}} 76 | ``` 77 | 78 | 79 | --- 80 | Complete codes and documentations will comming soon : ) 81 | -------------------------------------------------------------------------------- /test_predict.ipynb: -------------------------------------------------------------------------------- 1 | { 2 | "cells": [ 3 | { 4 | "cell_type": "code", 5 | "execution_count": 1, 6 | "metadata": { 7 | "scrolled": true 8 | }, 9 | "outputs": [ 10 | { 11 | "name": "stderr", 12 | "output_type": "stream", 13 | "text": [ 14 | "Using TensorFlow backend.\n", 15 | "WARNING: Logging before flag parsing goes to stderr.\n", 16 | "W0316 00:46:05.693278 139799807493952 deprecation_wrapper.py:119] From /home/amir/envs/Tensorflow/lib/python3.6/site-packages/keras/backend/tensorflow_backend.py:517: The name tf.placeholder is deprecated. Please use tf.compat.v1.placeholder instead.\n", 17 | "\n", 18 | "W0316 00:46:05.825887 139799807493952 deprecation_wrapper.py:119] From /home/amir/envs/Tensorflow/lib/python3.6/site-packages/keras/backend/tensorflow_backend.py:4138: The name tf.random_uniform is deprecated. Please use tf.random.uniform instead.\n", 19 | "\n", 20 | "W0316 00:46:05.852330 139799807493952 deprecation_wrapper.py:119] From /home/amir/envs/Tensorflow/lib/python3.6/site-packages/keras/backend/tensorflow_backend.py:245: The name tf.get_default_graph is deprecated. Please use tf.compat.v1.get_default_graph instead.\n", 21 | "\n", 22 | "W0316 00:46:05.853332 139799807493952 deprecation_wrapper.py:119] From /home/amir/envs/Tensorflow/lib/python3.6/site-packages/keras/backend/tensorflow_backend.py:174: The name tf.get_default_session is deprecated. Please use tf.compat.v1.get_default_session instead.\n", 23 | "\n", 24 | "W0316 00:46:05.854082 139799807493952 deprecation_wrapper.py:119] From /home/amir/envs/Tensorflow/lib/python3.6/site-packages/keras/backend/tensorflow_backend.py:181: The name tf.ConfigProto is deprecated. Please use tf.compat.v1.ConfigProto instead.\n", 25 | "\n", 26 | "W0316 00:46:05.911045 139799807493952 deprecation_wrapper.py:119] From /home/amir/envs/Tensorflow/lib/python3.6/site-packages/keras/backend/tensorflow_backend.py:1834: The name tf.nn.fused_batch_norm is deprecated. Please use tf.compat.v1.nn.fused_batch_norm instead.\n", 27 | "\n", 28 | "W0316 00:46:06.127520 139799807493952 deprecation_wrapper.py:119] From /home/amir/envs/Tensorflow/lib/python3.6/site-packages/keras/backend/tensorflow_backend.py:3976: The name tf.nn.max_pool is deprecated. Please use tf.nn.max_pool2d instead.\n", 29 | "\n", 30 | "W0316 00:46:06.494869 139799807493952 deprecation_wrapper.py:119] From /home/amir/envs/Tensorflow/lib/python3.6/site-packages/keras/backend/tensorflow_backend.py:3980: The name tf.nn.avg_pool is deprecated. Please use tf.nn.avg_pool2d instead.\n", 31 | "\n", 32 | "W0316 00:46:21.479355 139799807493952 deprecation_wrapper.py:119] From /home/amir/envs/Tensorflow/lib/python3.6/site-packages/keras/optimizers.py:790: The name tf.train.Optimizer is deprecated. Please use tf.compat.v1.train.Optimizer instead.\n", 33 | "\n", 34 | "W0316 00:46:21.596255 139799807493952 deprecation.py:323] From /home/amir/envs/Tensorflow/lib/python3.6/site-packages/tensorflow/python/ops/math_grad.py:1250: add_dispatch_support..wrapper (from tensorflow.python.ops.array_ops) is deprecated and will be removed in a future version.\n", 35 | "Instructions for updating:\n", 36 | "Use tf.where in 2.0, which has the same broadcast rule as np.where\n" 37 | ] 38 | } 39 | ], 40 | "source": [ 41 | "from keras.models import load_model\n", 42 | "model = load_model('CarModel_final_7.h5')" 43 | ] 44 | }, 45 | { 46 | "cell_type": "code", 47 | "execution_count": 2, 48 | "metadata": {}, 49 | "outputs": [], 50 | "source": [ 51 | "from keras.preprocessing import image\n", 52 | "from matplotlib.pyplot import imshow\n", 53 | "import numpy as np\n", 54 | "%matplotlib inline" 55 | ] 56 | }, 57 | { 58 | "cell_type": "code", 59 | "execution_count": 3, 60 | "metadata": {}, 61 | "outputs": [], 62 | "source": [ 63 | "def predict_classes(predict):\n", 64 | " index = np.where(predict[0] == max(predict[0]))[0][0]\n", 65 | " name_dic = {0:\"206\", 1:\"206 SD\", 2:\"207\", 3:\"405\", 4:\"arisan\", 5:\"dena\", 6:\"mazda-2000\", 7:\"neysan\", 8:\"pars\", 9:\"peykan\", 10:\"pride\", 11:\"rana\", 12:\"rio\", 13:\"samand\", 14:\"sayna\", 15:\"soren\", 16:\"suzuki\", 17:\"tiba\", 18:\"tondar90\"}\n", 66 | " return name_dic[index]" 67 | ] 68 | }, 69 | { 70 | "cell_type": "code", 71 | "execution_count": 4, 72 | "metadata": {}, 73 | "outputs": [ 74 | { 75 | "data": { 76 | "image/png": "iVBORw0KGgoAAAANSUhEUgAAAQEAAAD8CAYAAAB3lxGOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAgAElEQVR4nOy9WbAlx3nf+cvaznL32/uKbgANgMROABRpU5QcMmlJVki2ZoKy9WDPjGL44PG79TATipiHCb/O08TowWOPw2vI4/AStDY6aMmySJHD4QYSBEAs3Q30fvsuZ6slM+chl8qqU+d2gyCsjujzRXSfe+pUZWVmVX75/3ahtWZJS1rSw0vRn3cHlrSkJf350pIJLGlJDzktmcCSlvSQ05IJLGlJDzktmcCSlvSQ05IJLGlJDzl9ZExACPHzQogfCiHeFEL85kd1nyUtaUkfjsRH4ScghIiB14HPAVeBrwN/U2v9/Z/4zZa0pCV9KPqokMAngTe11m9prQvgnwO/8hHda0lLWtKHoOQjavcMcCX4fhX4qUUnZ1mmB8OB/x5HEVJKtNZEkeFTSinzoxBEQjS+CyEAmEM1AgSieUgIBILYtuuuiKPYt6XtUSEEWoNt3rcvAFlVREKAPTadzgCoKokW+H+AH0OSpQBsbW4BkMYxWZqRJInvj+93FNn7a39vNxYRnDc34HuQH1twrkbXExG0U4+72cIHud8HJ81cd1rknrf5u6svmnmAq+f+dHMRRRFaaZSSlGUJwGg0Ik1TpJQA7O/vApAkCUpJxuMxVVmZ6+PIPD/7/kQi8v00fa3fUaWl/01r5Ttj3ssIpRRKKzsi+xwic1wqiVK2z0IQxbF/OO49iaLYXKOVH7EQpv3ROL+ttT7WnpmPignck4QQXwS+CBDHEduba34gWZZRFAVaa+LYDEraidFaI6KIXq/n2kFKSZqm9Pt9APKq9PdJ09Q85Mq0nUQxWZZR2Yc9HKya9qWkN+iTZRm9nmFISmsGgwH9QYYWAqnMNTGaO7dvE0lNaRf/aHdiPkcjlIBCaFY21gC4dusmn/1Ln+Xn/+ovA7AyGALwyU+8SD/rMRwOkbY/ZVlCFJnxC/w44zhGS4XWgiRJ7FxIPyeeKQIijvwLGB738yeEn1d/XHYwUQLGJ+LGdzf37Wtc++73RRSe1z6ulEJr3Wg7/NuNP7yHX2RK+evrRUjj7zQ1zNjx3dl4Qp7nrK+vM5tOAXj33Xf50r//dzz//PMA/JN/+g9NPygRQnP7zi0218y7k2UZV65cYXNjgzRNqSrDHGSlmUymTCYTvxHMipntuUKqEqXMucPhgLWVLaSUFEUBwPr6OgBVVTEajxlNJxyMJn48w9UV0jhxEwRAL0kRIoLK3QeGwyErqwO+9AfffLfrWXxU4sB7wLng+1l7zJPW+re11i9rrV+OoqWRYklL+vOijwoJfB24JIS4iFn8fwP49YVnS8X0zl2Ug9ZxhEhiRBST5zkAg9UV0+EkYTqZkM8MpyvLkiiOWV1d9btFbGH3dDpl59ZtpNQ1x8RwVrfDSsu1i6Kg1+uxOhj6e/azHsNenzhL0UKhhEUq/ZRZkZPGCZEyu0sxMdeM90esrq6SH+yjpubYL/7cz/EXP/PTbKgMgCfOPw7Aam+FSivyvCRKzE4bZRERmjTtEQcQVymFciKMHWeamjFEreNa4HdSt+E7RFCLF81HEMc1I9YNiBk1dlFHbRHM3S8U4RYhghpd3D8SCM/Pssx/70Ii7l/YfnjMzYWD+sPhkPX1dfb39317Fy5c4OTpM7z++usAPPvsswC89fbrvP/+VZRSvv3JZMJwOPTXu/ajKAIdEcfxHCJz1ztUMxgM50Rf9zmZTJhMZszyovEci6JAmFfdi5NlWRqRRMnGeyGrxQLWR8IEtNaVEOLvAr8HxMA/0Fq/uuh8oTVJUSJcp6MYYUFKYSc6n0z9d4n20D9JEj+Zw6GB2e7h6kqC1kQa/9aLKDLigF38rh2lKvpJTD4eIS0cm4xGSBEjdUVZlkhMu3E/QaKJs5Q0M9evZwb6K5kTqQErScZLL70CwP/4t36Du3u7/PBbPwDgr/2y0ZGOZjOyRDAtC7yKQ8Hq2qqBqEVOEpsXPo4NS9CRILFw1jGwqqq8DsWNMbI6BUeufbdm2gtQ0g29tZaHwvrwGnede1EXLX73dxuuLzq3Tf75BgvePU8n6rTFISmlFxXcvaQ0IphOJevr66Yte49er8ev//qv81u/9b8A8PnP/SwAjz1+kS996d9z+fI73LhxC4Dd3R2OHDmCQqItAwMYj6eGKYrE6xpmdvPSQhlR1TLpoqjIpwdUVcV4PAYgn5X2t4JSVkhVP4vIPmMnOrjZi7RZEwmaJIlsP8bs7+8vnM+PTCegtf4S8KWPqv0lLWlJPxn6c1MMhhQLwWacUllEWgBlVUKWcWR7G6BWiGhY6SXMLGSXWiHiiMl0jMZw/TI3HLQqSiJpIC3S/JYXBUVZ4nSnd3fvALC5ukqZT9Czkn7s4CwMkAZRAIlVjpXTggLFbDJDpqZf8XEjrqwOByil+PSnP83f+h9+A4Dbt2/z5rtv88orxkCSTwz3LqoKVSl0JCgL0+c0itnd3QepiEXkFXhpmiK1Ji9L8sLcM7LsXwjR2Pk1ZjeWUnq46XZPB7VVsGMBKPs7dEP89rH23130QZCAG2dblHC/dZ3rzm8rOkNxwltmksTD8hp2G/Swt7dHVVVsbW0xtYrBnZ0dTpw4wU//9E8DMBqZ3fm5555jNpvwla98hfGBsRj0+32KYsbx48eZTaZ+18+yzD6bhKkVF93OrbW1RFmENhnPkKWiqioiu6+7vrjnmyUxMhAVoiiiqurnatqtx+qoKCqPQLrogWACaRRzdHWV3MrcIyWp0oStkye5eOkSAJmF3b00Y3UwZN8+AIlGpDEK7S0ITk7PJ1NGBweURf2Cz4qc/cmYwlkQhPltureHHE/o9xISa4bpJylDLciSjK2sz5qVRWVZkPR7HFBRZGayr07MJK+sb3Dy9Dn++t/4ArnV3v/HP/0vXHrqSZ56+WUA9i0DE0lMnPboJRFyYl6yspREaJIkoaoqdm7cMP2ezVCYh+10Fu6zKAqKoqgXOtpDX/cieY21lEgpvQgBdjHLGiY3Fis10/DndsjtoZxfm6tqPUNbPm8fd8yqLcaEv7WZQBzHBvomib+XsxJlWdZYDHEcE8cxg8HAn3v69EnALFYtBOPxmLI0c3Lm7Fl+9KO3+YVf+KsA/J2/80UA7ty5zUsvv0hVVfzZV/+L7990OkZrTb+fIex7vLo6pN/vE8cpE/t+OCaQ57mZQzvUPC/9MxgMjHWqCua8UhLiyDOBoigauoa+s5YpTZqm9NLMz18UJdYicpcuWqrll7Skh5weCCQggFRERoUIjGRBv7/KxsaGV9xtbR0BIBMxmxsbPPboBcA44OSqpEKT2F25mhlu3k9Shr0+gtrKMMqn5LKixOyMq0PDQX/v3/07dt+/hjwYc3TVKPlSAfFkxppIubC2zdHEnFvuH7C6ucEOBXct1696BqKfuXSJn/krv8jmyWP8zr/5twDslzlnn3icd3ZvAzAeHQAw3R8RCU2e50wORgDMZhPyqiTPc/KyaOywqjI7uOP+Dna2SYtaHGhrmt1OHSoStdYQKJ26dmonarVFhPvxEzhMsRgqug47p8tnwN0rCvxGHAJySMWhiDRNSdO0oRR2O/aGte8XRcHx4wYdnDt9hqtXr3JwsAfArVtGCfif/viPOX/hEZ597gXef/99AL72tT9lZdDn4GCPY0c3WVkxouEsnzIdjxFCUOTmnv49HI2sRarvxxjHMVIqJmOD3tbWzHtYKonCiL4heoMaWWxubJh2lAZlkJobe5ZlaJ0unN8HgwkIQZJESKvNrKaSqsq5dvsmb12z7gWlNYsRIauK4dBMntSKSktIYuLUXL/WN5OnlfI6AWd5KFHkskLbkW9tGIePQSIAjSwLZG6gW5YmICWD3oCjwxWO24kclwcMcsneaJ+9idG6nnnhKcDIjD/1mb/AP/md/4fvvPEaACtbW3znjdd47ep1c/1tC8vKkpSIcjpBWu+zUlZkgwwtjM7APWSlFEkUGXnR6jeIa5NSY5GIeuG0TYNtpxpPgUmpzTgaC1TML/r29/aC7jrHN+c19fMa//Z5bXEgHMNkMvF9dedqrf2icwxDK7AOff7a3d1dKyIp3n/fPKMvf/nLPHbhIl/72tcAOHPGuL3s7d3l29/+Dr/2a1/gc5/7HADf+MY3uHPnDnk+JUtjVtcGvv08n1GWFdqKqFmwMEGwubkJwGAwoCxLJgcjr8nf2jKepbLSIEFWtcjW6/Xo9/v+XDcnQmhm+ZQqL721DKEP1QksxYElLekhpwcCCWgBOk6orB1eJxFFWTFMU9askkSVhmsf29xmejBiMDDwbzweUyIpZUWSmeFcvXoVgHOnTrO/v4+U2jsibZ84xt07NxiuG7Tw1ltvmk6UBX0Jm1lCJa0GN4kZrAxJRcrt2zeZjs3xx46dptAlaZpy5uwpAFaeehKAv/7f/je8v7tLfzjk4x//OABFHHHr1i3SvrUK7Bno349jCgWxVjVEjQSTyYRKSUrZtNHLskRV0kA+jGIR8PDe7/pi3pW4TXPOP7o+1/WlrcEHiGLRQAwysCq04X/orLIIDXRZEBZZHLoclNzx8G9nQ+9qN44Sj2bc7jkajdBWQ+fGnsQp12/e8r76K6vGhTdJEibjGVeuXOWR848A8Gu/9jf5x//4/6Lf77O7e5fCugYfO7LNIOtx+fIVdu+aHfvs2QsAbMxm7Ozu+fv1ej0GgwGyKH3fb968CcDmkW2klGRZxnTfIJ71jQ1Go5FXIjrRcHIwIhYRW+ubfr4WiY2OHggmILVmdzZhYpnApKqMd1wS0V8xkGaYmsE+eu4RqrwgCmDp8ZPHuXb7podPM6uJnRyMePzRx7h69Spxaob63o3rPPuJ53jvxjUA3nzdnLt/Z0Kc9KikJrJ6AhEnCBFTCZigKa1J6YackAz7TKIUZcWJT//MzwCQy4rvffe73Lp2ncS+ZIlIqKY5WNEhs4s3H02II/OQEotRq6pAWu2+1Kq5SFRrcZT1wm1Af0EnpHbntUkI0WACoQjh72/nW+kmE2jELLSYQNt5KOxDm8EsihPw/evwDuwiNw8LGaCWc/1SgTlR2GNxnFAUhR/rdOr0CBGzWcEbr/+IY8eOAnDp8Sf55Cc/xR995ctMJjNWVgb+3Ol0TIRgOHDH6jGsrQxrp7iiMGbhOGbDyvfhGIqiYDVLvbdkURTMZjP/vZwZsSdJEiKMpcPFSNzLlLsUB5a0pIecHggkUEnJnemE0nLJMhJUUnEwHqPtbjq12vONlSGDNPNWgyzLkGjef/99z7WHfYMeSiW5vXOHvCw4c9xEUL5/8wbDIGzZKU9GdyMkmlxJZlZJV+QVVZoy0xF7RYWoDLfdHd8l1j3GiebsMaMwuvS88S3/wRtvceXKFfKDMcN1w9FjFTEuZqSVVWxZK8Z0NqUQkOdT7/kjpfETAMOhw51StiD0Iu6umLffH7bTQlPf17hGzEPvw+4food2TEO7/UW+A75PrQjA+3VfXoRO2lGVRVmLPQ4JuFBcEFRaeXu7izrt9zNmecnrb/6Ixy89CsClS4/zl//y57n+/lV2du6wsW7eqc31VfbuJsQipqqsAnNskef+PlmWsb1t3pHZbGb8DAY9ZrOmJl/EMePxqCHiTCYTtNZ+t3dixWAwAKWRReXPjePYK0O7aIkElrSkh5weCCSghKBKMqRlSTrSSCWZTKYMbOx9z0YB5lXJ+vo6R48fBwznq6qKLO0hLbdNrZx0/vx59vb22Dq6Tdo3xx559AJlWXL8qHFHfu/yW4DZKcpKIojYt2a5VESUUhEr4zMghdkNbpQHzO7eZevsaT7x/NMAXNszHox/8vWvMx5NkUWJtLoJUk2qNQOXoMIqBnUEhcyptPK7UiIitDCfgjppCVqDRQPOM5JAjg93CSfjtpOShJ9tWrTHatVtruv6+176hzYaCft8GBKIoqipn7gHhR6LYRtCmEwvXicgm9GExhU37EdEFZvfoiS1fYqpqoKD/YofvvYGYEx5584+wmd/+md57YffZ+eOUejJSpMkGcPhqm9RyfoPIbRXcAuhvVesQytO6actghFxxIH1LI2iiOPHj3tfgpl1a9ZaU1kX9MoGSCHn9TIhPRBMQEQRvY01elZhpgRM8lnzZbMvzGhi3DNdpNVgMEBKyXA49BPUSMogNNN8xs67xjYfZzGj8ZijVjw4ODCOO1ILkjRBRAnSKuDiJKXQJotLFQsKO1vJoIcUivMvPM+nPm9sxX/4R/8ZgP/81a/SjwdoqThxwvR5YHPYlJZJ5blxBikjKHVFhfbZZCQKLbSBpUIYTuFJo7SsNfLBwg5feOda2sUEFj6D4O+2Vr/NBO6l2Q99EcLz2m0sEifafXeRf/fL0Lr65r/rQxSXcYS0q1RVEVkSeZjtEsHkZY6Win6/z9tvmxwda2tr/MxnP8MLL3yC8XjMwZ5xMFKypN8b0ssGfmFOe0ak3N7eZjgceiZQFDPKMqfX65FZK5dzgNrZ3SXLMrJ+zyca0cDjjz/uz8nHxmpw9+5dJqMxhRDeNyDP80MZ6FIcWNKSHnJ6IJCAVIpRIcki0504FsRxiqxK704rrDJvdGATN1S1N1tRFBw7doLRyJw73jec+NixYxALyrIk65m2D0YjBsMel6+8A4TecaAiQS5rd0tEjEhN4EmEorQ6o7UjRzh77hyf/cuf4+KjTwDwP//W/wbAm2+8Q1VVbK5tsn7MpnNTJh9BYTNAqMzw3snkAB3VUA+gtIpBGSXEtMxisYHFTgwIIV5or3cmwnsp3kJqOgIGooU6RIRo7c5d9nvoSHrS2uHDc9tthfeq+9T0glzUN+czAGZ+hBBEIu70jNSRsOZBhxgkpRBENhmH6plzZ7MZQmt6vR57ewZFXr36PlJqto8e59y5c9zdMe7hO7dvESHIpxMfyOWQZ9ZL6PVTJlPzzu7u7aBkxJEjR7zbcWWD3K5df4/hYJXV9TWOHDniBsjGxoYfw4p1P45FRLG6hixLdneNiDqdTkmSiDcv73TO1QPBBBSwNxqDlWviRBhbulA+xHVr02rabTSY0+JqrZmMZ1SyTothPTRJ+z1msymlLNhaNe6Z+6M9hBC8955xRx72DXMZ6wnT3GRl6VtmVOrKQG1ZoZViy9pvT546x6c//Rd54ekXufojk0/11e/9EICD2QSlBVUSc31imFFPTkAq1tasf7ftaKFLYmWSSKY2TFlJ4zKllKJAe0cPrU12pDiuNcdCzC9AwEem6Q7ou3BB6275vGY0TdDoGEDYbrio3Tnt88P+hhr/w0SQNlO5bxEnYALuXnGkG27H9T3NfCkrDiRxZCI6bSSocz9WCiI00+mMylqLdnZ2+Pa3v82TT17i/PkL3L1lmMCt6zcYTcaM9g/YsyJCbNX0WT9lOh1z964RUw/GB2xuHCNJEm/7d5tRr9dDI9nbu8sjR4wuK04SHz0KoKy4gdJkWQ9hc2u6ue33M776zTc752kpDixpSQ85PRBIIIpi4qxHWRrOGllFV1Up9saGg2Y2b6DC7CBVWaduTpKM3d1dbzN1u+edO3cYrAwYj8fcvm24s4s9X7eRgqNRHXgSpSlaG/8CAK0FZWTcdAWK8xYJnDpxkmeefJqja1t86U/+AwBjC/NyrVk7skUhNFdvXrd9N6nPxIZNfzYy0DCNBChFGkVoaRRQcSxQSqKFoKgks8LtQIosTuilJpDInuznsIEGotrDrksh1Am39eH2+xp1NP0E3O8uKvF+kYD77PIyDKMAm33o9k84zHehHQS1SBmpdcvrMtJIWXlPTvdOmd3Z5Glw05/nJV/92te4cOE8Fy5cYPe2iTh87bXvU8xy1tbW/P3X142loFQFV69eprDPd21tje3tTQaDnu+DCy567LHHmEwmvH/tWiOAqKoqr7h837rK9zOTS6HMZ0EE7ob3QuyiB4IJ9LKUSxfPe2hjBldQVZU3gdRRZhDFMf2sTprgEmrUcrGZ8L39EXt7e2itmY7NZMcChNidh6GVRqvKh3QCTPMcqQr6/SFHjhzl8ceM/P/xF1/miRc+wRtXrvCvfu/3AeivmAmf3r1LJnOmBzm7QQTb8ePHOT1c8+MD0EoghSDpDXA911KgdUSe50b+tP1M4wyUQKmI0uL9NBCJQtldaGHqFhDAfLegzInNOgcAYh5mRzaXfngMu1C0UkThIo6iwMLQ0r5rkwgzChyiwpyAYCwx9gYorahk7for/K2bCziK58WJMHoyzCIUih8uCNP1V0Cdj9I1Lgsi8Jaiwm46sTIp64mFl9nv7O2Sqw2+/uprrJ04zdGLJo7k2MW3mFSvE1U5fSt2RqV5xw/2Jpw9cY6jWydM/wRUsWI8nniRz4kFWZyRkyNKjbbxK3HUI88nvo5B5FzKlWR35w5pHLFq9WiT0Zid23dYRA8EE3DZVFZXDZc0yUN1gzO3zw8fbpglB5ovhKB5vfPGa7zo1EqqPM/x2R8BEWt6vQFHjx7jZZsZ6Pnnn+fOnTt85St/xNtvvw3UqaA2NzfJZyWyqrwcmaV97u7s+fjztVXDlVdWVtDamHLcoojjtBEOG9ZXcPZyR6G+IJyn0K7eleU2nMeQ2iYzpRS0PPXCe7V1ADXNHwvvG2YlaocSh/2ozXrN7/VnMzQ6vL4LBXXpLRbRInNkGKhVM6CEPM955+3LvPHGG7zw7DMAvPzyy9y8eoV337hCav0PBmmdLi6OY49eTa2Iiirt+R38yBGjWK4qxc2bN9k72G/0PU1T3rfJTsNEpnmec+LYUd9OWZb+Pl201AksaUkPOT0QSKAoCq5cudLIAhPHTX/xumqMCRMNc8e1/crDHSMSzfTbDgl4GBok8kyShPG0jszq9/sUVcHq6iqPPvqYzz1/5swZ/uQ//ylf+tKXuHLFWBniyPTv2NETSCnZ2dnxbZdlyWQyo9/f8eeAgXvTaU6e50jpdvjKZoIxu5mbkzqxZO0THmq32zttO5S2S7O+KPlH26lnkV7hg1C4A3c5HoVIYK7ugO5OTx52a5FHXBeKWeTk1B7rIhNs/e6ZZ57nOYlS3Llzh+9+97s8cvYMABcvXuTxxy9x98Z1qtxVqDKRpEVREE0nTK14EEUROo3Z2931z+XEcROmvr29zaOPPkokYh9jsrN7l4ODA3buGph/9KiJaJxMJmxu1tmNwLzbhz2vB4IJSCnZ3d1tPGQh6tRR7hjUTCA81zGCNow0TKDpTQcmdr79QvZ6PYhi730IZvKm+ZQLFy5y/NhJVm1M+fVrN3n11Ve5deuOT3u2uWl+W1tbYTwe25DSFX+PPC+DBW36UZam9p1S9Uvm0oe5c8PYfkddirK24s21twj6housfU7bBbl9r3Y7bbofpuEWdVsE6VIMOp3AnPzfMa4uZhG2G/ajyywZnh8yhZAhVFVlagZYpaEsShCCycGI965e4/XXjSnuuWc/xlNPPcXt6+9z7cpl0yf7XPKyIC9LZrPC32/tyEYjpZ7z+FNJwunTp9HKmCMBLz67c1z6s42NDba3tymKGaMdY3I/fvw45bTePNq0FAeWtKSHnB4IJAA0IL5DAiGFO1EIzYC57+HuorVuJCBxv7d3NrfzJkldLaaqKkQcsbW1xbPPPusRwte/8U3ef/86p06d8oo7p9Tc27vLu+9eoaqURw6nTp2iyEsmU2NGvH7dmA4HgxUP1cJcgg7dpGk6l+rbVMVtKqYWKfgWUXvnc7tne6cFfEWdNmw//B7zFYjafy/034c5JIDq9n4UujvMuMuEGf7dFdPQ3vkPG5+r5+DqFbkCunfv3mWwuuKf78VHznPy9ClOnjzN1XeMAnnTJr7RmPiYiU0GUlQl4/GYiBhrDfft9Ho9ttY3KcvSx8w4cc7FErhsWrv7+0it6aVpXcJvsOJDmbvogWACTsPfhKeqIQ6ElWjrc5ruoe5YKDdrrUE3U2AtEjPcIgtf9AuPPMLLL32Sp576ONeum+iwy5evMh5POHH8pNdV1OWuNHGcomRF31Y3TpOMLB2Q2sCQgwPzIPf29uj1ejbLbA3f0zRt9NEdb8vsbTl4EVNoz7Vr617QPjS3hUxg0fV1+936iFAP455Xlx7CfffvQ+t3v4ipn1vbRNhuo+seXfdbdH74bjoTpzNriijyz08WJe++awKLTp04zisvvsBjlx7nO9/6JgCvfu/b5jxZEaeJtyqVskIKkGUVJEc1997c3GRn523KsvRux1VVcfPmTU5Z/cOm9SS8evUG16/f4Lw9DvDGG2/4iMQuWooDS1rSQ04fCgkIId4BDgAJVFrrl4UQ28C/AC4A7wBf0Fp3lz6p2zHwqrHDqYZdvK0ga8eiL/Iaayu/Iudx2HJWcbtLJevilmtra3zixZd59tlnGY1GfOtb3wLg5s1bSCnpDVa8OLC7axQ2/d6AkydOsbOz67XH+/sjsqxPv2/OPbJ91F6zS1VKyqKuJZCmKWjhnZ+cpcKNscueHo6jPa+Ldvf29YtokdKwywfhXteGSMB5braDeUJxb5GzUN3ufKXhRdaM8LdFCKbrHl3HpTQKXVc1WisTmJQN+hwcjLyy7uj2EZ549CKPP3aJp2zS2d/5l/8MMBmtTpw4QWTzZCQIVlfXiBDeQc55uZZlyd6eiXlxETKbR7bpDQdedN225fqceLmytkbPeRNev07ZEplD+kmIA39Ja307+P6bwJe11n9fCPGb9vvfO6yBLjkfmjC07aASlqZyD9fpFNqyboMpmKRdtSecP25kryRJPMQ/d+4cr7zyCo8//gRvvf0ur75qqgrnuakW7MQHc0/T3tGjxxmNJtbM51761MJo026oRyiKgv39fa8TiOMYrU3gUFVVQfval50KdSeOumTmRZr7+10IodNSl6jR0OA3Fk334g9FuyzLiOPYM9x2huMGg9MLFrf3JeqG8G2Rqr1ZdI3/XlaQ8F2trNgTpxlRZALBRqMR29tG7t/b2+PNN9/iheef5qVXPgnAmXPnAY9FG0AAACAASURBVBgMTG4AVy9hd3eX3d1dyrJkZ9fsmfu7xmX++PHjbG9vk2UZ77zzDlCnI3O6KAf3B4MBWmu2trb9BhJFxnr2ztVvdI77oxAHfgX4R/bvfwT8tY/gHkta0pJ+QvRhkYAGfl8YVf7/qbX+beCE1vqa/f06cOJ+GqqqqrWzzduP69/EnIa3/bfvoL2uXfXWUbs812g04uzZswD8yq/8Co8++iiz2Yyvfe1rATyrII4QIvYBSG5339/f94Uvw6CTNOkF4ah1GTGlFIPBwHPyPM/ZP9gjSRJb7bbwbUgp55SFboxtG/giaqOkcB7cb05ZFUWmcpML34babdYhssMQiPttZWWl0YYb9zz6M/d0ooIvuDrLvTIxbD+8R9tnoi3mHSZOubG3fU1CZbVDh+EzqGxeB2H9OkpZMRwOyaem31evvs+RzS0uXXqM4yet449NU37nzh1md3a8T0ClYbZ7QBRFjA/MO5XYe06mOVlvQJIKTp464/vR6/V8Do3+wPikZL0BRVGwu7/n+7yxtXmo2/CHZQKf0Vq/J4Q4DvyBEOK18EettRZtW58lIcQXgS/avztq0TUX76IXvFNr3HoxQ6ahWok6Qll8VpT0ej2ef/55wOQoPHbsGFfff5/bt2/PFdpwiw9qOBtFkZfpHINQSrG5FiOEgWdTayo0MRLGGuBe+CRtekKGL6rTkSyC+4dZBxad20WLnG3Ccc/L57Ujl8v1EMazuzG0r2nD7y7LQyiWdI3lw1I4n139WXSNS/HomKIsnROa+WE8mnDt2jWuX7/JI+dPA/Dscy8A8MPXvs/e3p53KOv1eoYBWssQ0DAHOtgfMsLQrF5bp0w0ZxylCFvcU1YaWRULx/+hxAGt9Xv28ybwr4FPAjeEEKcA7OfNBdf+ttb6ZatM/DDdWNKSlvQh6MdGAkKIFSDSWh/Yvz8P/K/AvwX+NvD37ee/+XHan9PqL6h1f48+mrZEd1kqqDmoi88+ffYcf/GnPwvAhUcfI+1lfP/73+fmzZsBtHQ2XInWTSTglIVZliGE4eTlLOeAA7LKIAGHKFwkmRDCV/2Npc00ZGMZQljbRh/3M/4PsmsuQhLtnTkM021HL4YK2jDuIRx32EaXktN9D5/9YTty+/p7jXmRUrSrD/fyLXBjkVJ7Hwo37qIouHbtBq+9/kM2Ngw6fOmTrwAmzdjd3X1frTjLMiM+itql3SkNy9JUqVZKUdi8E0qZZKf5pJm2zKMSNGF8zWFVnz+MOHAC+Ne2wwnwT7XWvyuE+DrwL4UQvwG8C3zhfhpzEwjzGuj2sbYsuujl9edZ7WhIbV0BGLPNSy+9xDPPmFDQra0tLl++zA9+8APG47E3+UmpUDSdVBwEVEpa3/+ILVt8ZDKZMJ2OmZXmobpgDydfSlU1wj61NkUl2ousa4zt8bT/Puy8RW21F1Sb8bSZghl3vdCdOODaaVsAwr/ddW1rT5cX6P2Oq+u7O9a1sLvEgXtROCfm+siKtUkDoo+mE3705tucPm1Knp9/5CIAW0ePoJRi78AEFK0OV5hMZmgxX0ItSTKUMOnyEp35tkPHNscwVGn0G6UqG+LYR8IEtNZvAc93HL8D/NyP2+6SlrSk/7r0QLkNh3bktjIp3EU+kDgQxQjRjKBSRHNKK5mXnD97ms9+9rOcPn3a3/Nb3/oW169fR+tas1wp7XfpLoWb1ppYCKKeg8N9s7vZ/EGNsYmmj4Nrw+2MoQvzYbA4pA+rOGv6VTRDk0Mfgq7+hDt2Ex3Mn9sF4T2clc36Ct0I5oOHOS9CCB8EBcy3FVk/jtQ+/7odWSlu3rnNlfeMwezS4wYJnD1zns3to9y8bhLNlNKEkIdFaH0WuSy1Ima9XJ146LT+NUqdL7gaJkLpoqXb8JKW9JDTA4MEwt3G7IxNc9j9yMLtY4YrN39TRLjUZeG5/X6fjz/zHC+99JK32V+5coXXX38dKSVxlNbKPxF5PwZpSz1JVaONJI0pZzkTq7TRWjMY9hB2t3emnzRLQNJwmU4SJ1saL8R2+rNF1NaRtOftg5gIw7nRzOtg3K7ezuvQlrmd2bPtA2LmTTb6FAaILVJ+zo/h3nJ+1zi7ko/ej05hrj3dPO5MuHXaN0kUwWQy4+ZNYyRbXTNRf09+7CneeustJtbOP5lMyOyzd55+k4n5TZUFSZIwmegGMjSfTX8IR1Ece7/NqNnVOXogmICj8OWbyy7Tovu1iYs5hUjUUJQ4GP7xj3+cz3/+8xw7cYqDPeO2+eqrr3Lnzh3SNLb9sZF+sRMl8CGa7uE0YXLtlBxFEbF90cvAZlv31SlAa6WSGX9zkWmtcYUFwmC9w6D2B6VG+DJ1pJ5rt22xCMmJdlDPiTsn9PloO+6E8+d+D2H/Yc5QXVD+fpheW5l6P4zR9Ss8PxxXyMCEML9NZzOf9MM5fz3yyCM899xzXL1sko1cvnwZrUGqWgzylTR0nUOwqOrktVVV+USyc8p00R0a3kVLcWBJS3rI6YFAAm3I6rh+yP3vB551Q8CWOTGah7YAzzzzDJ/61KeIooi7trzUG2+8QZ7nDAYDcludyFFb2VWLMhpQ9PuZdyV2+eFdqi6XA342m6GUoqoKr+BxO0ztmtusLWD+sPf8ALbxrvMW7Z5tcSBEZKF40s5cXKOV5j1dfgQHcx2CcO7DbepKAxbuvPX7Mm8+bo/xsPHd73VdJITw6cVcYJdSijh435IkpddLmO5PfVDQ7bsm2vTcqVNcuHCBkyeN6XB3dxdVKcaTiUcLkTDLM4ptqrFAEejfD+sVmAXvj5QS1TFfi+iBYAKgiSPp4XZZFBC53IFNt2FlIZLLci+icLC+trn9buy2+wf7bG0Z27yBSoo8N+GeL75orJy//Mu/xJFjR7lz+y4/sCWnd/YmkK1ydzwjixMKZSP9hCZJYrSqX5o4qu3cADJ8oaKYOI0QbnwW2jnRQeu4sXC0NmKBcyRy4/f/fNp0/D3bi/teupTG7FsRw41HBxKkwNQpcL9Vqg7BFo32NWgX2OfC+8xnMctJ0x4iNc9lZbDqmUlZmDlx8QpVVaEd03Q+BKkgSVJSe72r6ixEhIiw700YO6B9jYNQGhYCNIEDVjBHkYuNCKJVoWa0HqJL6Re6sveYTsekaUrai6hkRSXsuaqimpj3WEvT9/07ZoMZr25y8fwFXnrxEwC89eaPmOYHCFWSpXbzkMp/9rKhrZfg+qXpZRnai1F2Y9CKNEqQNF3OD2NwDwQTcLtEcyFoFPXL50169ncv54iOAQon0xrf6o3NzSBbi1HSOY76sz/7s4BxDHI+/GEhx9I6/1S6Vm4ppSiKgiSKvULLLWyn8OqSSbt2PadQCr8vyhkQngNNR6fw+CJloDvmdu2GHCmDpJqt59DV1uFMxj4rEYYOJ4E5SxDHJk1anAR5FKiVpFVVIZWZU6UqH8NvvtvnizKZLFq1JbSWc3PgxitVHQDkc/53ICqHWtwv4dwmIvLmZzMew6yLqkRKSeqchQSUeUG/X1cVcrUnnnv2aYrphEceeQSAJ554gu9951s24a0riOsyKFmlcdR+NxQ6WAcAkXZMXflNsV2vok1LncCSlvSQ0wODBEITkRDGEaeUFdqa4kTAr0yIq9uhormwVg/JK7O7ra2ucuumyc/uQlodXP/MZz4DwPr6Onmes7e3x3vvmaSNRVEgkpiNjQ2qvGgkIHX6AJdFJrx3GHYLNRxrR0TOJdQ8hO7lANOFPg5zEGnv4EpqXAqfUM/RjuC7H3K3ddKZm3MdiEMOMblztXWkEpEwJcviBOcfludVwzrjEr4q7Zxg2laK+boLblya2hzbjgo9zPLRFLfsPFkg5mR4ETdNhC5Z7HQ6xeW5dEig1+uBrDwS+NjHPsaVd99m584ueoFZM8zdaPoa1SKhs1x5s2fYRnzou/BAMIFICPpZ1gB1SmmipO58W+YNJ6QNTX3O/qi+1sG7lZUVlK68V+C5c+cAOHLkCCJOWFlZ8Q+x3+8zKwum0ynj/QPvPyCEKVemk3Qu0ajz/W+b74yH4byI0GXau5dJp62kc/fvMnnd7+IVYr48mfm8v2ClpiKvGUqcpHEjGMq06ZR97p7uGTpxIfYec0JkKFUxmxmuMLb2cyf/O8WcaztkXm3GWFaF74eH/IHo0I5VaXuWgoHc5k7NuIeqqtBB4I4R7SoG/T5J0hRBisIog53fwDPPPMOPXn+DW3f+DC2b77y2JtKIQBns9UN28YO/J0AW1cxOa40Si9+ppTiwpCU95PRAIAEhBMPhsNYYS2k15MJbA8Lim+6a8NP8bT7dTqQzs6tWlfKJPtbX15lMRz5S0JEWpnLuYDDwJaGLokTu75FlGWowqGGoUkRRUzsuvSY79vA3zIuolGokDXWfH8Qs1Xag6srME85JW3HYprYZbh41KGBxevFF7TlUUVW1A1DDpKjnHVkcpNZae9QQjlMIwXBl0BiX1vPl0MNU9eZ4E0JHNt8j1O+U+y0MufVm2qAylOtvEicNZZtEIytrsotDR7SIXi/zbYBJVOPGFMd19qStrW2efeF5vvv9V31EoNveI63nzKTgHMfsvQLrSFs8DJ3OuuiBYAJRFPkUVABVIEM72J+XdXGOts14LidR8IL1e0P290f0BiZUd3VlneOnjvNzP2cCHbdsMYg0TVlbXeP8+fP86q/+KgCj0ZhpkZNlGflkyv6+SfzoHpJWyr9IPhVWUXiZN2QCoawYetS5f+EDK8tyDsq6RbTId+JeOoH2uYuCn+yVtOmDmBzb15gF2RTtmgsVr4+ofwO3CnpRk3k6V20h4kBnoefuGUVJg9kYXVIdnFZbK+I5MSxJEpPizfp6uA1ACOEXnPftiAT9fp/NzU0Gg0HDtbfX65GmsSl9Djz+2EV/Tx1Fvl0pJZcuXeLixYu8/vrrAORFzaTMM2uG24tIk2onegR6Jl0hq/lArkW0FAeWtKSHnB4YJDAc9gMLQYRswbyeNBzZiAeauKPKzZx9PIoY9FdI0x6DFcNx11Y3ePq5Z3nllVfsOeaag4MDiqIin5WcOWOSOSqlEVY5ubW+QbBRIKViNp122qPbu3KtcGo6onTFR4SIIWw33L27bPahUqt5Txr3dL+10YcQMXVQlGy0G7bV5bnXfg5d1Z3aXprOCay9g7v7hUrJKMpsrYKmki5EAO24hMju+A3EGEVEcZ1wM1QC+t+DcHattYfyIcLI4mQONeRVaSoNt/L+KVUZK4EVUYvcBZUpZrMZ2xtG9NzZ2WFra5uXXn6Fq+8ZC0JeBOU6lERo7T3ERARCicCiY+cmBqFjrzgM52sRPUBMYNgwE0qpKYOY8jStPfK6MqV0vXxSmuwr6+vrrKyZRI0b61u8+OKLPProo+YcXWt407RHHKU+qWOeF0Rpwv7+PmVZeHOgq00QeqrVOd4j7+0WLoLDHHe65sP9Hp7bpi7LQvjpFntX26E44BjXopclFE0O18k485qLqFxc9dfJ4O5YHDcDYJpz08zdEFoCuqwpjgmE5/prO/QR4e/hPYQQc1l6HUNsM1itFGnUNIUqWZm8Ero2gTqxUQjN6ODAZ5lKen10VfLxj3+cP/zDPwTqaFOEhlzbd6sI5joQDV2uCltTI8lC86ZYqD8y1yxpSUt6qOmBQAIiEiRRXNtHBYAmiUDYIIoweWXogtsfZHPcWVaGO66trYGImc0Kv7u//PLLPProo41CkADD4aopMhnHfsfXGqqioN/vN6oBCSGoZNHQvrfLmoU7ZDsGPDynS8nXxbXvRyF3P+d0td3eYdvttHfwELK3RRvze/MecZzO7dZt5WRb7BFRXd0J3UQCdSBSRZfSy8HwtgOZ63c7a5UTjdoipVKqGQOCTdrZMT9uZ1ZBXok4jkEopCyJo2b1qbLMOXr0KDdu3ABMgNBKf8Dq6iqf//znAfid3/kdwPhFlGXJYNCrrSn5lOFw6EVHl9hf6To1fT12zYLM/8ADwgTAwhmvKI7AwrzE2/2S1rnBS1OaiXc+21Fmfa2TmGmes7m55T2zHnvsMba2tpDSnOMSfDpYqVWYdCKmGYDStcjaL2ENYT8IzekzDr3n4ut/3PPMwp0XQxZd39XPeX1Fs71FVg2oxYFOUUZ314pow/L6+u40GoaRdBeu6Rpnm/mFfe42p9rfCTwRTQa5Rr4E85mgtPBuhyJKUALirMfJk3V6O3e/rY1144wU1XUyjZmxKdp06WwOE/XgAWECob3dHNAIbarf1JNWR3NppT03jrQxKUZRROoKd9jnEUURSRRz9uxpnn32WQCe+NhTrK+vsz8y0Vyr2Yptt0v+XsAARC2Tdv4+95vTCbhjovUZNC2a9+16meevEQvPu3/mEraxWFYOFYpthZ9bOGH482G0SE8CoCQo6Zhxc6HXyMrIu7pRer5WcIZ9MwzD7JQ+OrGV4bhrcbd9LkJG1zapCve3rsdiQnq1RwguF6DbrduKyCzLOPuI8SVwG9Qsn9Lr9ciLqe9Xr5eyvz+vmK6ZovY9utcGsdQJLGlJDzk9EEgAIBFRnU4JGnA//HRyW8NpQmiiqObWVVWH9Z44cYInnniCj33sY4BxDnL+/YacRcHJtFETWjkY1QENBV3oAJ/oob6Hsjv8Txbaf5DrPphOoTk3bc18LfvrhkVmUYGXNmLwjoMBOnJTpZRGSo1LGBK2PW99cLt41Hof5tGPEQOkT5Iy/3t3rEXbktDMjRiYGH1f2gjBzFE71Zr7DKtaF8WMNE45etSg0yeeeAKAH3z/VaPDksqLvEhFpINYgra4EmQtNnkeHnRxQAjiWBBZF0iJMylJpA+6qWGUgVJOxqorrLThZZZlPPPMMzz11FPe9i+lNEkds2Y1oNDNN3zgYR9rCqC96JIvFejmi79IvvxJ0P0s/sPk/HYboQkuzHHoPrvOPSxox1wTzy0ex3QhLEZS+eNtG389lsXmyS7z6iL6IHMyPx7h4/yFP+bG47hcsPBaykjfhktGE0lmswkq1t4s+clPmnLmo4N9vve975AlkU87XlWVKVBjdQK1AjO2TKCZWOYwJrAUB5a0pIecHgwkQDPttiDIsechpTk3jgVRJFCq9h5LksSotSyXdXXfzp49xwsvvMD58+e9kmV3f4RSilUbFuzIm1Vopvn2MKtrdzN/zI9HGE8ureZ3qoVz8BNEB/fT1txuHdX9bSKguLH7p61cdocpHN2ObZR38yJFl1lUa2OVOUybHXlLUROFuPt3Ke/MO9NEbd1iQ0uE6RRH3fPvVrz5bx3a+bKq+6xbRXDiOKaqKm/2u3TpEgCX332Hd999m2I28feoqsqKzNa6ECizhaCR+g7gkEjiJRJY0pIednowkIBoupAaBhw3OLQIOGb4mWV1qKbjyM4l+Omnn+XixYv0+sPAXVM0XJS1Mz26kla6uRuEximXzahpTpsfS/vv+5FPF7Vxv9d+ECTRpRhDa18ZuZnzsHufWIQAQhOhaOlL/DXazHNDD+zyEsZ1VJ93+VUmZsTrwHyXjPzdNeft707hqzvOOcxU6aip+7DoYm5utNmcPXg0uqNwTmq/iJiyqs2VaZrS6/UYjUYeCbjw9yefvMTtm+/zve99z+vCVleGtkhus/irH1krb+WH8hMQQvwD4JeAm1rrZ+yxbeBfABeAd4AvaK3vCtOT/x34RWAC/Hda62/e6x6mz7ViLwJ03IShIlD4hPbckAm4CXEx20888QQbGxtMZ4X3w+4NVuj3+z5mwF1TFAWJaCpv7hV4EVoEPir6cRnJIupkALgcCS4TUBqcLxrWgTDMue2Y0mW5CT8j5z8QNQu2QD3XJlAoCVKHBV6EzmbfUP7VQUTuWNi3esHb7E4dHoOL5jZUcobneNEjdBBy5yFbUNz1p/mumBwLlb+u1+vR6w3MeyrqBKZgCpWMD17g6tXL5FPjzbqyOqSqKvrDQWMsyvabVmzIYZvE/YgD/xD4+dax3wS+rLW+BHzZfgf4BeCS/fdF4P+4j/aXtKQl/TnSPZGA1vqPhBAXWod/BfhZ+/c/Ar4C/D17/P/WhgV9VQixKYQ4pbW+dug9BFQJhPgw9rnTLNwpzfcs7YEwlYHBhnGiycsZ/b7hik8+/RwAw81NikgzFSXKuaXGJUma+viCxNucnc+1rJWAQvk+qRYjNZy1mfw0HIHWdX0AEdWQFMAmkEep2j4d06xbQNCO/9vf3Hq86VqU0To81+yaWuDzhzfMbeaPhuJSA1o6r0CbxcmhEIHHmVFiI9IEREEWHW1Ts5vds/T3cuMWQtTJ+bSJDhVRfX/pouMigwLCugMiTkhsARMIshApYRXKtZLOmcpEnKAB6b0JpbHYqdpY5nZaj/j0fA3ENhKTSmG82kVTHPB+IE1TqEMlPs7EVshWkSDtpxSF+T7JZ6hEwkrKe/umZNmxI9sArPaP8cilp3jllVt886tfBWB0Z4/jW0fYn9pIw4F9DhHkZUmviH0thTC8uYt+XJ3AiWBhXwdO2L/PAFeC867aY3NMQAjxRQxaYDDsk4hmeLA2b7V/b0QA4ZTC1wPMZzNKrRgMVjh7xvhcr6+b5I1Z2qMqSqudtROCAKXnEjO6e2qt/YIXVvRvw9z2dSF1u5UacuIGSSsjrKrTP7X9vMPFq3F6ichfByb91Jxbsk1/5kzVUStTbQROsQw6atYaaGvPRZMZNfU3jrkKIleUo6UvcXPnrDeCiKg1fbELFY8ik+MhEv6ZJUncCju2HY/rcfm5brWr230JQokbUL8F+9uiYF3nohtaO0eh7uPzBXDb1hH3PU1T0tjMReLyElSK1dVVnn76aW5fvw7A97/3HZSqmUtZOuaiiKPI+8H4ex6SXehDWwfsrv+BBVat9W9rrV/WWr/c62X3vmBJS1rSR0I/LhK44WC+EOIUcNMefw84F5x31h47nLTRqkYEkMUGemjPwUxXpayAyCcZyWclSRZzZGubJx5/EoC1FaNVzdKU8XiMiBLSgVW2iMikgPew0SqbbELNLtjfxfnbyqRF7s1dv9Vtu/x4TVfaRQorrWVzl1Vt3usip4Tf8ep7z+8EorUTeQpt7qLj91Y/zSXz9Qm6UBY03ZDDY2CUha52YwhhzfF55VrVqjlRozhBV3WqWMcNdNXufzsxTedYqMW4cHz3Q+E70UYCMTFZ0qOfmXBjl41oVs1I05THLl3i1g2z1K5cuUJRzuogKJeNSkEvSxvPPrpH/35cJvBvgb8N/H37+W+C439XCPHPgZ8C9u6lDwDr86Dq19t8dxDGPtzA1TTJ6vROURSxMljl9InTnD9jrAL+hVCK8f6Y4dqqh+KRjoxM7iYliFBTSqGDgqUKTUS3htj3vfWydC0GL2a0Frf1KWo8sLaeoW5H4R6XX5gy9K/XuNwLc1i73Xc1P57wCn/PaH6x+DlSTWehMFNveyG1x98VkdfO1OMyNkPtX19bbRJ/XEpJHKV+8Tu2YcSmlplZmH7PWVyUYZJNvUrbwhGIqkqh0Wgdytlqwd/3Mhs7M6odr4Rhz2a1rlqiQxTzyMULAFx68glef/11ytKYvnuxdYPXmhiBLIMCuvewDtyPifCfYZSAR4UQV4Hfwiz+fymE+A3gXeAL9vQvYcyDb2JMhP/9vdpf0pKW9OdL92Md+JsLfvq5jnM18D998G4IkzTRV8U1iuSYIGgkdrtgRJJljSpBR44c4fTJU2zYai6ltSRoImazGf2VoUlUAqhKUqmyVvS4ewq708SRVzpS64uave1QFHZF0LVFA+dkEkL0tr7G7Oiu3cDzJLyuJQY4tYxPLa9cpqWqYwcIAnOc0lU0RQfnQ6Gr2tbudlrjym2VXS234TZEDo93JQBpQ2JH3mU8QHRa6yAxR+u1FbV7uRfpJHVxzlbfvN/CgqhHqGsddJ0rhLAK60Up3eeRQJc41ZwvBSpCaOilZlcfH5gU98Nhnxy4s3uX1SMmRf7TL77Itds3uXPTZCaKfYnyuoRfHDtkFn84JPBfg4SoNaGAgWyRiTzzTCB1GuEIRERlodLG+ionT5/h6PETpD0TH1C46rXCloZSdaJFZWsFDGyaJ2duiuPYVEGWwsNgrbVPULJIFOiSL9sPOIywc98BK2q0XkwNUF+rWiHV4d+BwnquDya8er6gSN3fpskoFpHXN6iqok2+uEUL2nY5C7UpdOBpj6VNQtTeo20Y7cSC9oJsLzJhmaTS7SzJuvFO+YzVITP3Nl7Dfv0SbScWXagTuv/UcO1nWpam7Hls3z/n4LaxsQFxRLFzm15mRIWzFx7hzCPn2d/fB2BmP4WOiUVMlmWd72YXLWMHlrSkh5weECRg04g7xhjXLsQ+WZKoPxUVqbWDbm5tcfz0KVbW1vwu5ZKHyjKisjkDHRKoqoqqKKgs8nAxBUmWIbVqWAcUmkjO23XbO5s71j4vHF9IsZjfyVwuhTYEDZVtOure3dvXETVRSLsP5u+ScA8QYnHFmvY956FsDfkPQwK1JblbBIA6bgDUXD4H17ZzZqoqcyw8b/45NZO8ai29GFSKuvSZUsojiHDM7fG4XBcQ+D655y0Oj2Nw7YbXhIrCUknjL+PyhthxjaYTRBaRrQ3JrROdTASXnnqS29ZacNkigaqq6KcZVRpU+Z6zIjXpgWACflGJujsOyrnFXxU2PiCLUQhWVo134NrGOpvbG8S91J+bW615TIlCo9BezpVSUkpNz743vqy0EEhtznMZjrSWRlfR0d92OGyYFNL9C0UQKaXX5ofnCj2fdfewslHGGcdQaELTuq5Q6zzEGiJIEGjSyRiYf2mNmU3MiTJh0sz2S98l/x4Gh91vYTnvKDI1+tx9EjeWKPHXQV0iXnc4AIk6/e5c/xy5sOgkYLRhzr928ZLwUwvRMLGG4lA4V+35aDOB8Lt5PVQjRybA7t4eyUpGb2XI3p4pSDIZjzhz/hxHTxwH4Opbb5k5KUt0r9fQGykOTzS6FAeW617zqgAAIABJREFUtKSHnB4IJAAGIFaq3mGcki8vZ/YMt3tGZGnCUaslffTRCzUHttrQSta212mRoyNBbvMOVmVBmmReDIgCe7AqS6RSJFlq76ioCmNvHQwGjQpEcRwxnU7nElVEIjL15LXym5CqJFopjzpCxxhUs8ioUoqqqmw66ThIu2ViG4qiqHcRqygbjfdNZGQAi40Nv1kUFTtiM4/zu3VXtaL2Lt6lsHOfi5SQXd9ri0kzlZtrP9y53K7sqJcZFFgpE4MwHA5ry4aURoGoav8Fdz+fgKYlKg37/bnsv2tra2RZ5p/5ysqKb68sjXXJJarRWjOdTn0RUzfXrixZqKRz/ZnNZo1xVlWFIiISEcr2/a4tgHv24jlGxZQk7bO2YSxg+XSMlJJPvWJSkH3vz75u5yZDKVMB27kOHxwcHIouHwwmYBewsvJ3pSRRy8zS9w4jFVGl2RgYLemJ7S16gxW0rKhsFdcsc3ULayjkKgknIqKQM++i58wx0/EEIkF/MAigeUw2NKWsy7KcCzgJsyGFi9jByjA8Vmtd+3MHTMA53PjkmdYJxy3KaW5ewqIoSNOM4XDoC6c42XjY6zOdTIMQ2coyEOk9LiO/WK25TjY156Fo42vpOY871a3/mPtbuLiEbvjfhsmdDka+rZox5kVJmqY1E12xYlVpysxVVdMUGosIkRgY7sYSC21MtMHCcxaisA9ZYjeAokSW1RwTLYoCpRRpmjKzmX6EMElCV1eP1qXDwOet2NnZ8c++nebcMRn3jmg0ldVpheXalVJMZ2OE3Sh7SUo8K8jt9Vs298Du7i5JmjIpimbbhxgIHggmIDAdrWz0mdYCLQRKV16Js24X9nRcMNrbZ8sqPgZaM4wT7u7tka6bkQ5WDbd87/o1MltQ1KfFqqSXOYE6WCNJiJIYoTVjyzDSNEWkNeeuK9/MR4eFO1pbNvaKtBbDcLt1uMjaqGDLvtyz2YxKSbIso99vlupWSlHMcq9Qcouii/t3KTUjkWAq5ZgXzL18Yb9Cm71/bgvMhV2/h99DFNCW5R0SaCgdBWRp389x2qvHL7VhJmE2336ws2dJMxty2K7byZUpSGCCd4L0aeGnLK0pOYkZDodmYcmaQWxtbFIVOavDFd/+bDajIGdjbd333S3MyPbZK0RFZCIUKVG2jPnYIoHZbEqaCHq9DG3f2+FgSBb3QJi5WI/N+tidzuj3e5RCMBqNzHylw8bzbtNSJ7CkJT3k9EAgAY1GCe1leqEjk1ylAm1hclLYdGKzkuLmbb7++/8RgDPrR/jY88+zcfQ4O/umqlBkZartlTV2D/YRsqKyosbQ5hwoLUd1MneWGT3B/v4+Ywu3+/0+sd1Ry7JsVE2GetcBA/3AlELzu2c7NLpFcRwjWmJP2zrg7rm6uuq14S5U+uplE7X9xhtv0OunDU24Qz9tM5YSzKGYtnNO34pI5rswbpP2egeX2zul9v/VkD6OmnuMP9d5ANJEB/48c6A+FkdMxgd+l3afJ06eMIhLxDV6wWTpEZFGBKG2IRJoiz2yNOJTlmV1QlAr97vvDjnGUUwaJ8biFJid0zT1zyesMiSlJE3TuTJk4Xy45yW0QklFatHLwL5bqYCiKin3C1bs/DMtWZEaOTHjPnjPxulNpiaPRpZQWeSqpZyrqRjSA8EEoAWD0URSglQkztw3MgszORiTjWbcvP4GAH/2u7/Pjbcv8/zLP8Xx0yafgBvw7nhMOZ4iZ4WXs0phFDX+BXDQazxmZ/cuk8mE1MqKo9GIqiiQUtLr9YIy6amH2860Fb6gbTv6nEdbIJ+LlvwsIuMtadJPGTkQDBweDocN0ea9q1cB+L3f/d2GnT+LjfIrTiIvn7trQhdS/xlFJEHdhVBBGNMsBb/I669rrIsyAIciSag3gdqc2khpZjMVO4b7uV/4RQDOnj0LkaCqKrZshulSGYWqsK7ToRnVMYK2iCOSxDNCB9edbO8UggObnbosS79xzPKJ7/NsZ8bJkye5evUqzhvzxAnDpN59910/zu3tbT83TT8LSSyM7O68MyfW9l+Mp/T6KQmafmH6PhAJan+Pb3z5PwFw63VjItw4cRS5d8BsNSWNawbmXeE7aCkOLGlJDzk9MEgApVDOCURVaGIyrX1MdTw2XLc/KzlCxJGVTQDe+Nr/y/e+9k3e+u4P+dRnPwvA6b/waQDu3LrJxtYmgzSjZ6HRdDZFaLzyxWvIhYHeq6urbB05ApgdcXxwYMtCDzwk7PV6XrnYdrxxO80i7zkIdlEV5L1xSUu1CezRUpmS7c5fPjb9mU6nXkG5tWXMpGfOnGJyMPJjqcoSIYyCT9tx5tTFLJVqOigJIfzuBkHaLURDCeqOuTGEGuf7UQy2j7d3ad9uSzmoYqP829szirK+RWoOkYWIItEmN4Xu6JPQTd9+XzcgBqUr8lnuLS+j/QOEEEymRrnmSttD7VTkUKCZUzVntdnf30drbXbilgK0Douur49FhNASSutANTGoZLWXsbG+hpzO2H3XoL9r777Hrdfe5Nt//KcADHIzrq2kRznJmYrSi42xiFiUNRoeICagtfQdFUIYbW0UM7BvmhrZbMFFxTBbYWTl/3UdsV9Jbr39Lv/q7X8MwLO24nBvMOTipcc5efIkkZWlellGUcx8JFqocU7TFKm1h4BRFNXlzi08h/rl7fV6czbnEEov0qS3yXnmmYvMAihLYxZzfZFSMpvNDPS0kPWJJx4HYGv9C4xGI591Oc9zdFWR57lnXMXMfJZlaf/V5i+3oGXZMk3Jpg8DGF2KY3JVEBnYZWJ0NAe/A5u9ozA4KEmShqY+yhLSfs+3e+GcyRuRxDFKa59XwM157XoczLE2opYKxkLQHwfznUu3m3cnFuxWu76fjrk48cSN+c6dO6ytrfnxHhyM0Vpz9OjROfGpnpPgoFKoSnronlnLVDmZ8e7NG7zz2mvcffMdAN771qvsvXmFbeszcXbDbFzVwYyVjQG9NEZbZtRLY6qOpDKOluLAkpb0kNMDggQMV3S7QYJAVZokikmdo0Buode0YJBmbB83PtM3dvZIZjkRtab/T//4TwB46523+aVf/ev0hwNO2d1DC2Prdbn1Qw1vFBmwe3BgkIRxlFFe8+t2prC+fZ3osfSfXdldw+8eUtvvDSTgYoAs8nD9m05NLfqNzTWmE6epNudeuHCB2WzmtfFpmngxIG7BQAefnUjgxtmwSgQZip1/Q7ibdykHjZbe2v+5N7X97N3chr97dBUbx6m0b8SA3CKV6XQKQtDr9Sgq52OiG0pG30ctQRolqB8L9Vy76wa92tpTlqUXvZynXlVVFEVBr9drJPMcjUbEWXMMg8GANE0bWZLC0HWttU8Q6i011J6b3/nWtwG4/M7b3Lr+Hm9++7tE+waZPLa6xbljJ9i0/gE3dm8D8N7tW2zEx1g5edwjxrSXoYr58HBHSySwpCU95PRAIAFTBDQhDkqTCyUppWRiTXsbWPlmOmV7JsluGYXNNjEqW2UyViibm+1bu3cAGKwP+P4f/Aeu/X/f4PEnTa33Jx9/kmeefpaeVfREx4xy7W6ZE6+vMs1zXOo4rTW9JKY3SNnbOyB2ZaCl9DoFhwDcrtWzSiuz24SJKGpOrHxeOavv0FUjulZrbRKgCqMkBUh7ZpfJyxyRWAWqc7OuCkjqkmmlrOptjlaSC3efiGALEKgg7blL9V2Po6ZQwdVONNo+p02LIvIA/5xDpaBX3CkzF9PcKTftvWIziLysAn0SQUJS0RowECQarZyXoVLGTBrVc3gwMfK8Qx9FoARM0hRNjUiEiEkzE7uR9XvekzAv94mFZGU4ZDw20X/CKiZ7cZ9qljObmDGdPnWa2zd32Ep7HLxnsgXd/g9/YOZmJSNNNM9Jybrt4fndfR6JVol3jbL0mtX9PDZc5603r3L35FEGPfOOj/ICLZooJaQHggkYX5T6pfr/2XuzGE2z9K7zd867fVssGZGRe1VmZWVtvVRX0wvtBWzZxmIQwgMSDEjD6sFcGCEkLgZ8M0iWJS4GI9+MJSNGgyUYwwwgLMaexoylBrvtsrvL7qW6qruWzKrMrKxcYv22dz1nLs7yLvF9EVnlLaSKJ5WKiHdfzvucZ/k//8fBiCMh0dbMm9nAVk9IMhSVcpwBOcNBHxGFhH1jyl1/9iIA4aDHt999h+39fX7rt74CwK033+Hl33iZq9euA/Dn/uZfBaAfxmilqfIcrGk2HA4pJ3OKNCOS3QIbRbOEtb4X7dcfvs92MKjedjGQYxFGvxmAU8fUiTflcSC9yzAA3e0XYQK6sN9F+zddii524rj9m9K+voplBm33HHbvpdsu2ncZ3LZ97cq7FM2mKQ4LkGWZz0ZVNoOwOVyljCCKzXHeeO11bt+8w4Nb7/DGK8YN0FaxXb54iWI+4UwgiffM5LdSRcQ6IVq3UPaxUTITkTMPNJP9A4Zn1sw6XRGEpziBUzmVU1kiJ8MSwBbTuNiYSdxSUBK5ydcGYSazA+JKcs4ixHRekUrNw+kUaa2D8xeMJXD1mac5d+0p7t5/wM62cRHKec7d2+/x1hsGYXX+hrEInv9jnyZggC5Kj9iSWqOqily5ajVLfCENRFerCmkhoupQCmb5LHJ4Nlo+oy9C2dXrDm/nZJmFcix2gcPMQs0AXnfmX7R82X0vsiYWnX/ZOWrpzl3LZus2M5FZpmg2MHXr2gSvy++jZan4oKMhGVEC8qrRPi0MCIRAV4rIzuqbq8b93Lv3gNs3b/HWG28C8Gtf+q8MowGyqtgcmbH9sc/8MQDWkpiH77yNnBdIiweYzgseqYw122NjMjLfx/2DMfNBxGw6pb9qsQ1aESYn3B0A0Aq0dC/fVMWJomDd+mRnLpoPO600Dx88Yt+aVZEMWds4RznsgcUC3N81plH+zjuIXo+ol3DhooEUkynOb5zj/j1Dy/Sf/+N/AmBtfYPLz16nH0UeU1DmBaKqbEmyaLAhlyAC68JYX7YxTps4/aa48uh2qrrptx4+RvfvZUrguP0Xre8O+sN9EdoAnvY1Lz7msg+nKYuUQVeJLHMp7BE7y5YDs7rHN/Dqw9e1iHinvb6GTTul6CaLSpc1m7XWPtbQSyK0rgiBUd9WF44N6O2/ffG/8O3XXiewp7hwZpPLZy8xPdjnrO1BmNh4znh3l+17D3l6dYNEu6pXjZIh90qTAXg/MOPwnswphwlJGJHYuFWpNIFsZ6uacuoOnMqpfMTlxFgCRupeABKBkJIoMpbA2S3T83S/qtgpCg72TS6/THP2Z/uUQiK00YarzmwvctZWV8inmvnEaOAYydVLV7j+xFMA/F//yTRPevV3f4eN85uMzm+S2trxsjCNSwMhqIrc4xgqZWwDUyVmVXmn60+FXmD1L+CjRx3lDfy+yyKz2/1+nPXQ3e/o2ZpDlsUii2ART8EyyPXh7MLilmKLZGG1YiOIe3SQl0PLpZStDtRVZVrECSmohKsYTNBFQYBgY8UE6f7Vz/8CAK//zu+yMlxl84yZ9be2tigmGfPJmB3LI5DbYN700Q4owZmVNTJbNXiQjpnHMfvWWrgbmzG/1484c+Uc62pIZF3VQFWII1hFTowScFBhMBDPfpIQiwFBbEEW9iNLzm5waTREW5DFwwfb7B6M2Z9MPKvq2RXjC52/coV4OCQcDNgOjPmf7c94tLfLas+kEy+fvwDAq7/zNV767GdY39qkmFtGliS27aZsPYC9VqUqcB14rd/pzEHfXXjBM+9SkTkFsGgAP44JvWzbDyOL0n2L3IFlWYRFZvyizMFxsii2sCy16PoINJd1pb4HRzrSTH/WRCTHZU+6166UIvDvWhuqMwGBFG0QVlESxQnYPhm//Wu/BsAz15/mEy98glu3bgHw4MEDHt67jwolsSXR1QPj5xcBjDbWKRDsWjdYaQh7ETNLsDJzsOuVhPjcFoNJwKysyWFcdmKRnLoDp3IqH3E5MZYA1AU6Qtd175UFfkwsMGMwHCKSHtJq5ljA1sUtooMJmYVGPrDVZtW77zIrMpL+0BON5Crn4b2HjGwDx82tLQDe/cbXmB6MKbMcHI9bHDLNcsoyI4pjT4RqeDbMLO6DTN2Iuj4cGPQbu+j0ggAVHA3NbR/vaBP4g0q3wm1Zrrw56x9lsTxOcHKRPG41ormO0l7H4mMsDsK2eQObFsZxWZCmFG5/2yfD0eT76sKkQiqNzkv23zcAIEdn9/TTT1NUFal1XYerI9YkTNI5+6WZ7cvUjOcyFChVMpaaed/sf/7pJyiSgIkl4K1SM+ZHwwG5FAzimL2x4SNQofDFZYvkcRqS/u/AnwUeaK0/YZf9Y+BvAw/tZj+htf4lu+4fAT+KQXH8Pa31F487B7Sj0yG2qqvK6LnOtBsmtTJOZ0wP9n2VVxpHyChkNpkwtxyF2UNj+uskJFWKfG/P8/slBPR6IYVDmLlKLVUxm46ZTg4IbDpFaii1Ii1y+qNhzQ8nJRpBpWvftUkO4gdOd4z7lmKt57vwWbify3x4s+DxlcBx/nvzg15aU2Cl+/E0j7GMOad7rq4CWeSzd5cdlTE5qq9g+/4WV3l2j9G9vkW8/U7xB14JGK5CZV1VXZmGJpUqOTgwIB9fsh1IXr/5plcCuztzpqpkfzph64pxUfOx2edA5/RCkwYst0xsYTuG9/YeEFvWbW2zaL3BgErhQUrm4uMjFfLjuAP/B/CnFyz/Z1rrl+x/pwA+Bvxl4ON2n/9NdJvencqpnMqJksfpSvxfhRDXHvN4PwL8gtY6A24KId4EPg/8xlE7OU3vcPfz8YSNlTVUUfLo/vsAnLMQyCAI0VHEQ8sZECUxZVFQhILKVnFpG7G/ffcuJCEzS1kNMEr69AkoQ+MOPHH5CgAyjvjaN77OjU9+zFf5FfOMMtAM1lY4mE99dqBUllFYYxqYAqFcPkv7GatTM9AkIOnOOstmxlY+Xx1hJSx5xq1LbMykTdqydhNPwe3bt/27KcuSvb09yrJkMBh4so3xeMyTTz7JmTNnfJNMR8mV57nJq1ur7ihg0AcJci7KRHwYWWTtuOM331X3nI7GvcgyZBQitUCXFbGd92YHYzZ7A/LxhIl1UR2Ap7++SjCM0YV7z/D6W9/m0uWL3NkxBnZsg3k5mjKEVx/eIXE8nPMcBjGTytYv2PlclKCykoNxys7ODgAbly+0CFC68nuJCfxdIcRfA74C/AOt9S5wGfjNxjZ37LJDIoT4MeDHAAb9Xsv86vV6xpTJc+OjA/fuGZ8q7kUEUYS0FOQlkOuKSivmjmfdlaUqgSpKyiz1oJ4gCNBB5HqZsDczJleQBOzs7TGZTBhGRuFQCspAoMqi4UkadJjAfOfKpoMqV3zj71A3iD8dFNJ9XK4/QO2bNn3PblvurvgPRR9WFIv+Xrgvh831Jre+k/F4zMsvv8yWjZ0Mh0NWV1cZDodEUeQH2nw+54tf/CIvvvgi165dA9rFVd2+gkfJsYpAdJ7rod+Pkcdwo7rZjkWui2qsU2WJ0gIphC/pDoqKLE0JgfffN5PZyhnDiLU7HbM9G5PaYqNKgh7FpFIzt1H9fObGsyQTAVmpkcJNJBpZlgQ2Ld4LLRFOpqjGKZO9CTv7hgglPrNCFS83yD+sEvhZ4CcxUZafBP4p8Lc+yAG01j8H/BzAxvqq1pWisCmbXhiQzqckBP5DuveuoVUK4oj1s+sMbX13Zb+6MAwRbja2TUiEEIQyoicllR0kqsjJVOUr8HZs4YUIAu7cucPDhw/pj8wDzYqcMkg8y4/rZ+jGoGjMIKG09fuYgKakmXbqIvDayqC5zs0+y/zhVjrrMQbzImWy6CMToqYYa1ZCOkajF154ATDkmfv7+9y6dcuwHNttn332WebzOdPptNWAA0w3n65F45q0HHVNx8kyZOGxItpKdOEmx6QcWxgHAWhtuk+VFT3HM1AaZqc4jLj5zk0A1s8aXMDde++xc7BPZWNThTY4g9lsRjW3TXILpzg186pgXrpuXJBISU+G9JW1rizHxFzl7D/aZe9g7Hkxzh/znD5UilBrfV9rXWkzmv85xuQHuAs80dj0il12KqdyKidUPpQlIIS4qLW+Z//888A37e+/CPxrIcRPA5eAZ4DfOu54WmuKou5CS2kua+PMOrNHxtw82DN+ZqUVWZaxakETyXBgGFyQ9K2fn9v2ULoqkapHoCoILINQUVDKgNBiqX17sjji3Tu3uXfvHhcumzqDWZaSRoLZbOb7FZj7b3QTcpToDWYYYwksoPZu8BTUIg/N/M3oedNXdss9jr3jKXwQMM4iF8KZ784iKMuSyWTCZDLx8ZCvfe1rvPzyy3z5y19Ga80TTxidf/ny5VZbsaY0+wy6v5dF+h9LRNc0X1Y78Hhz3CL6vWUZidY62d4mkpAWGTHGklRFQZmmyLDkjbffAuDppw1S9c2bb5kWc7ZOZTybUkYl44MJceGsVuueVTnjbE4hFaHNZsX9AXEQG2sZ2H9oCuSKtGDn0S7TquavWF1dJevGrBryOCnC/xP4fuCsEOIO8L8A3y+EeAlj594C/o59GK8KIf4t8C2M1/3julnKdYQ08+x5nrE+WuHyhYu896YxowJrvhd5zuThI7K55YVfWyUb9dFV7VfH7lhFiVAZsipdCAAVCAhDwoEJ0LiBP+wPKLKS9957n+esL5ZXJVWaU6U5BbI9IHTV8qODBT6p0G1zN4gWc/dLGS40M5vb+GM203C6bWJ3t2vuvwx11xR3XEebFccxKysrlGXJyy+/DMArr7zCO++8w3Q65dy5c9yxvQ9+9Vd/lfX1daqq8ulb92zTNG0RtWqtl5KRHiXdwKqRoyHGjbs7fLzme9CHn1dXCXeliQMJhaQKQ2aTKdpS4akshyLn0fiAN60SuHrVKM39RzvIMKDUxoxPDw4QfQlZSaTstaaWPLUsEDqnN4oJ7Ucv84oqS5nnZpuHd82cXJWa6XSO6kX0+sZVW19fY2c2PnT9/tqXrqkf1F9ZsPhfHLH9TwE/ddxxT+VUTuVkyIlADLr0kSPOLHPNYNDj7NYGVW40paN01BqyomK2a9yDIsuZ7AWUZe756M8NTXR/NYjo9QbkVcnMAjhmWqGRCDtL5VPjVqyvrNPv93nv9h2mFqTRXx2R5RWhllDUdeh+hhDKI39UZ+ZpAlPM7xJV1A1X6+Ug5eHAzVHm8gcJDNbnqFNcy6wBt+3enokqz+dz9vb22N/f59VXXwVMy7Oqqrhx4wbPPPMMDx+adNbu7i69Xo/9/X3vYrnjzWazVgfnbmegRfe7UA6hr9SRls1yWdCeXS8GLx26hMb5WvsHAWhNURTkgaVsL0wp+ptvvcWDbUMEmrv29gjKWUZlx0QwzaEQrEQJa9KM4yS07oCsGKuUUkM2NcfO0jHzaYqe2zZkO+Z7CHt930MjsS7symjEQV73nejKyVACCIJGm6QgMU0l+v1+7XPb5x1oUwnoW2bN55Qp5HlKZZVAbmuuZRgRiJAE2UDpCbQI8eEHmw6rypJ+nHDv3j0fVV1ZW0UVJZEMUGXlr1EpDcImpVymQC42x5viUcML0HOLXIBm9NthCpomqqoO9zFcJtUC2Gj3nO7DdH0Vi6JgMBjwsY99jBXb+np1dZU33niDg4MDNjY2eO655wCjOO7evcuZM2d82rCJC+hmJ7rtxP9wlYA9XBPHtkChNt9Ls6fhItdLYOJDoWi0bAsDlK544+23PB/h7rZ5NokIKGYT+jbWMugNyfOU1TBiWJrjxzb1FeuKPK3IJjNmNqVdZBnlLEVYd0BS95UUukJo7eM4/X6/xYzcldMColM5lY+4nAhLwImbrXq92PQBQBBZbZY74E1REUYBiS3EyClRKIPys2He+dTM5ONpziTaQYYRwloJot8jEAH9oZntlM3T5nmOlJLd3d1GOymTJw8TiaqUZ+FVuEmpUe7qWIfE4VmtsuaCI011TMVuppfycOFKk6wSjo5WOzkKf98EJjnpohTdOof4y/OcM2fO8NJLL7FpW7NdvnyZJEl4/fXX2dvb4xOf+ARgrIc7d+4wHA4PdWMKgqDd5NT+7bj3u7L8HrtB0g+OMGyb/PXybpuu7nGX1Uu4ZUprUyvQsN6UxVncvnvHZ1weuEIiBVWWs2ED1HGYMB8r4hKmlhlrZrkDCkr28xnjYsasMO6E0CCruk0flneDMKDIKqo8p2evOQpC4mD5p34ilIAWpsFEbBs/zCYzems9xlrx/Hf9cQD+86P/B4AgGZFOpkhrvoUippjN6YUDgsoVA5kX2hsNqJSil8RoSxQSTaaonV3/kkNbu/0wS8mqimhlhTffeA2Ap56+xmClz3g8ZrS64htc5GWFDAO0FkjR/kgBUAot6mUBZhBJ2yhi0UBaxOsHx6TVGh9F279d+JjtgRcvrlRZm+82Ct8f9FC6YvPsBuvrJs4yHA0oyhyE5ru/+7t8yurMxjrf9/1/EqUrlM2KOBdJdRJElXWnqgYN++NIt4Nz25B9nPhIe7vmh93lGHSbOe5Id27daMyiglphx2FEVRWsrgyYWABapDUHj7bZv/se685cf2DcAZmXXNCgbpr28mVVEmQFmarI7DjLLDpQSQEaegJCO4ayqjJIWVsNO53YWENasT4csVembK4ZYNJaMuT2tEaBduXUHTiVU/mIy8mwBGzgyEWVwbRnCsOQZ543gadf/L//PWCiqucvXGD7kYm2pmmGkNJoT6tBe5attSwVs+mYbCrohXaWC0OK+ZwVm8t2ZprWGq0UVZazZ4M3Wil0VZnmmw0+eaUUUknDOuSvePF9LbtfOL5E9nHlKJP6cQNmTVPZ/ZzP52RZ1mq8OhwOee6559jY2CCOY19nUJYlo9GIPM8PRf6PusZl17coUr/ouS1b90Fl0f5HvYdmgFbWZgPZbI7D5cRByPb9hxTzlMyWEu9ok03pa0FWKURR+WO4BqIehGQgtjoxAAAgAElEQVTHW5ZVTLOUtCzBjmPCkDjuUbpzW4tsxZbcb4xWOXvJlCT3Voa+ZHmRnAglAHYQWtM6iUJKrRjPZ1x/8ioAn/78ZwH4+ld/h4d7O57gI+4lqLIyRUe2m4vDrieh6eqbhBG4TrtK0esNvMLxhRUyQChNWRTcv/seYNKHKo68kjrkT+u6PXdruGjtW2P7bTn64z/uwz/Odz5O4RwnzbRdM6rsWrJvW8r2NE0py9IDgFzUOQiCViXih7mH4/Zb9MEfp1AfV47bf+E7dK6BlFCZ5zefzxha/1wqzVvf/g7lLPVgt9xyXYYiQGhB4ABUQKlLZFDXy2j7gfd6PcI4IsoLJjZlnhcFs7xA2s7I5y0IaZ7lPNx+xEuf/C6+94d+0FxHGHrK/kVy6g6cyql8xOVEWAIaKKrKh2y0gDhJyMqCg5mZ1X/4z/4ZAJJej9e/+SqlrRScjyeUKCa7O34GO2dz2kIIiqwgTcf0bMQ6RFMqRWAj9FK42S8gFBDJgINdA5aZTab0NteJgsD0t3dYhUU5/dYN6UPuwePOUB9kRv8glsQHEZcdcZ2Y4zj20f3JZOL5DoqiqGHTQXBkzfoy8/7DuCsfJFOy6PyLruOo8y07jp9BK4VGEEcRuVK+E3Q6mfLWa99GpAUbAzMmB5Z/ItEaqRWBa15T5ugoIm08Q2kts9l0TpplFFqjnKclApJBn3jVuLXvH5gxO1xb5cXv/jx/7q/8Ra49ZeoUvvjFLzJVy9/NqSVwKqfyEZcTYQkIUfuUYAp3okGP/mjoLYGrlgHoT/6pH+Azn/8cX/ylXwZgPp6x82gbwqmv405dAUdl0H5h0vPpqgCIQ4m05ypcqwMhjF8nJHNbm/3o/ftceeLSoVnBo/do5PVbM8SCGn4hvMb9MIHBowpZjprlHlfCMPSzvfPrhRDked4iOQnDkCiKGI1GjMdjf24XIGyiAxcVP3WXLwvoPc5M3ty2a2kcF9Q76lhum2MDu41rV1qZbtUaEvv8HuyNeXTvPrJUDIbGd+87mnNlAs6VNM+q1DCrKtMKqYOSTYYDktGIXFcUdtm0ytFJgkv8nX/qSQA+/YXP8eJLL7F5+Qo3Hxgik2B1xKPJ76GA6A9DtNIUeVUPHikYz1PSsmJtzYAp9mwgL1lZ5eyFi/xP14ypc+utt/nWN77Jt775qg/2uRqCMDBw36rIPUlooSArNZFt7CAsiKKSkkqYYIzLGLzx1pucf+kTBroc1HUCYHL0gjrrXA8MCxiiPcCd0nB/N/c56vfussfNBHxQ96AJ421G990y90yiKPKBV1fzAUY5OMDV7zXrcdxHfNS2H2TfRQHGRevdNs1l5n3a7IBlTsqyDHRd5fnw/gOqvCAWoe+ToWzpoRSSUjSyS1FIEZiel7gJyioDmURIEYIqKS3lWDJYQScJw3MGC/AX/vr/CMDGhXPEvT53th940NfWlUvszdtEL005dQdO5VQ+4nIiLIGiLLn/aNs3hYySGIVmmqecG5lGpNPUmOhJEDJ+9ICz60YDXn76GleuPskP/ukf5u5dQ2L0lf/2ZcBANGfTMVEvJnWpPK3M9G0tAF8THkjDCRBIQnsht969w/eU1aHCHcERQar2j3rdY85Qy8zQoyyB7jZOPsis6GC8UKdYy7IkSRIGg4FPBRZF4Wc+KaV3A3o2VVUUhaccW3ZdR8lR13zSUoTdZS5F6qi+375106RRe6HveFq5fWRAlcQUjiNBSlRPMi9KpHUnHFfho9mMqtSEvZizl8z3cOnak5y7eoUrNvh3/glD5TkrS8qqYGPrHMISlty8eZO8WE7rcSKUAJgH6Hj/8mJO0hsQ9XtUDt9tQRLJaEhRljzYN4CeUdIniXoIGXDFugjPPPssYNhg37tzl/2dXR69b0gXxnt77Dy4z8xxuo9NVLVEkFUlSkJoX8LOeH+hSSiEQKvFLbeabcVaA+oYH/1xsgLHmdcfxI/uShP440hBHGlKnuc1Xz6QZRn9fp88z/0HP5vNiOO4RZra/Pm4MYrHARB9mOzChzn3onffXOcyS1VVgTR8lmEYMrZj65337lBJiQhDpMUOFNYtqKhQImBmTX5FxaQqkaH0zW9WrGJ9+tlneeLaVdbPnPGsV1uXL9FfGVHYlNW+JSftDQcURcEsTX22ptfrLWR8cnLqDpzKqXzE5YRYAgJEQGjr9fMyQ9gCnb0DE9Xs28aL09kMEYdE0szWFYJ5kbXYfeeZnbWUpr+5wdr5s9z42PP2TAq08mQlE3v8/fGE3YN9Ki2ZpiZP/trr3+HNN9/k2tPXiaKI1M6UeZYRxhFlWbbqzO2dtIKATpRSh7ID3d+PfUq/D1HvZdu73gDuWt16Z967dffv3+fRo0ekaWpg3c88A5iuuuPxmDyvuzd7WK1sU7NJKU2V6JLZ6Th3YNG23Vn7qI5EHwZvsJC3wO0mA4Qwbchm8zmVne1v3n6XMIkoFeSRfbZ25hdhSBVKkpEpzOoPBjxxaYvV1VXOWXr35583Y3Y0XCGMDceGK2LLpCTNUqS1kN39FmmGEhAFMb3YWBIS6X9fJCdECWBKOW0aT8oQQYBqfEouiq9lAJWqU3ICQFsAj9m/cuBtKai0AQdlwkXvFVqVvpO4WDGm79bGOufCayACJg7aORoaEtMo8qkyc63CtKJuQG0XuQMnSY4b+IvIMqIo8vfq4gRvv/023/zmN5lMJly8eNG7DmEYIqVslRK3Iula/76Z8cvSikdt+zjbPU6moeUuNE5fVBqlSwSaAxuJf+bjL3D5zFm2RmsMbbXflYsGz99ftaa8bXmXDPtUUUCSJB4Sn9lUd7i6ipSS8WyKsGS6LgujnbJzbTK1Nl+O0p6QB6U5gmf0hCgBOwCdjyWl67AjDtVxa4FRDp4iVqK1QAHSBllKi9gKhKDU5mG4YgyBQiiFtk8tti8hq0oiBFEQEg1NefHa5gZ3br/HxtZZQ9nkcPJhUNcMLAniaY4fpL+fKL/HlWXnbM7Wzv8PguCQpeNm8Tw3aVeHLlRKkSTJoVn/D+Laj1ImywKoiz7y45TJUcdv/u4+SIGAQPDePRN/unbjGT7+zDNsra4T2QHjUtUqDgmjEGHHX64qpkWOaszug6FBGe7lme/T4M4ZaDPlhXbiE7q23qQWaK3Qrh15pXy3pEVyGhM4lVP5iMuJsAQEgKg564XlVtOVqlt5uZ/Cdsr1WtyU9AYIn37xvi0G/KOE8KgeSYCWAscvp20JZpbPSMuKKKoIbLxhsDJiPp8T9xK0EN49ycvCnENK3/LZ1RNoasRga9aAQ3GCkyRNVGBzmdbam/kA165dI45jnyFwfQdGoxFKmZ4Qy/zo32sab5EclVV53BRpc9vHKW3uLnP7BdLUT7x7xxCFXL96jeHGGrmGzEXq18xzJAzIdeWzRqWU9IYGGOfS1jNbH1NVFSIM6MkaVasRBAiUa3/nUWvuevDgtiad/yI5EUrAo6YaTlaAzdc77n4XUBESoTTaMcY66i4BiC6LbDOPLN0is481czMbxEmGI6Q2YC1lzajB0LAJlWVJ0u/7NmSlqlBA2KDkat2PxQU05aSbXM0PwMGGpZS+SMgF+zY2NhiNRoRh6PsSgEkbZlnWQht224w1A49/UHGTD4tSPC4msDBF6JR9pVAYYs8izXz8ZDAcQhgxTzM/Tny78EJQoQlcEFWYoGGv1yN01HmFcbVMvCWgqupSbaHNF9BVsEpolNJIodGOHbvx+yI56WPzVE7lVP6A5URYAkaaASVJEAREUeSLMeqthJ3x29x+TaJI2eD1NxpY4nsGSIFC+B4AlWVzCYKAvCyJg9hr7SgICYKARzs7nNncILJpStng0PcuTOMaBe2ComVykrIIi5qDuoxBVVWtAqIwDD1JqAOkZJmZ7Rxy0O0PLC0o+rBy3HE+iCuwaL9Ffy86lptcK2vZhDJgPp2xYs361dVVhJQUWjGwWaiqAbpKwthbTnmeE0UhZZ6jLLovsZ23q6oikEEL0CURxga2hpdLQGupUc5CcfciJUc5BCdDCejDLLhhENBP+vRiR85po6JBgJRha6CqDnovsIeqqPnYvauhbKTAOlEODiuEIJ+nDNcGhInle496XL161Z/LVdkpzPVWWtXUUh/ig/79Rrw97jnhsAJqsg03W4p3P+Tm300kodaaOI690oB2IVI3qv4HlUXo+vddmO/jZmwWQZS796Ecs5A26FShYX9vj9VV0zG71+uh0KZ/ZtEm+gxESFEUPo0XhxFpZgqwXD7PFR1JLVFlxTyd4ciFlU8RuliUM/0lWmqElD7LIALZSrd35dQdOJVT+YjL4zQkfQL4eeA8Job3c1rrnxFCbAD/BriGaUr6l7TWu8KozJ8B/gwwA/6G1vqV485jACXWtNZmFonjmCQyM7WbXcIgQKtawyu05+33JrpH7xnXoWoQgios5ls6em0z68VxzBRTGJOnloWlAqRkNpuxrs4QCPO4gqaW9UmLxfnp5v39UZr/y/LmThYF6xzIp9kzQGvTPSiO4xaS0BUUKaW8JdHN5/9BBgOPQ0Qel/NftO640mw/gwpDSF7mOdvb28QNPgZ3nGRgsCfziQn2SakJZd13IU9ToihuYTOa7dJ8psZhYKRABHW3o7JJ3x5Ic02uq5cUR3asexx3oAT+gdb6FSHECvBVIcSvAH8D+P+01v9ECPEPgX8I/M/Af4dpSf4M8MeBn7U/l4pGU5WpRz9prZFBhJAhWPqlzU0DpSzLkiyd+RfkG1s03AmHuDIvQHfowAI7uC25g0VyVXnF2tqaUQZzgxgMRMAzLzzHL//yL/Psx573/piqKsIwNNDXwBWR1GajDALChrnrPgCXqmma0Ie59I1icQOo6SY5P9yb22GNVnT/W8cRoka2NmMm1pRssidrpfzxHMTUQJ0l/aRfg4UQ6MimaSuFthmTQEiLWtNE9p05R1TYf+7vUIY2JVQrhuZ9dk3vRQjE9n1K/xz9q+4gN+vYEYeWLZIgaLsP3hV0/n8YImyX6TRNKfKCMAzZ2d/jc5/7nL19TZqm9Pt9j9hLLDhIa02pSqLY0ospKNFUVIgGZRsYmjEhArRS3rd3CqBQzp0w4ziKDJxdVRX9xCgeXWnkEUb/se6A1vqem8m11mPgNeAy8CPAv7Sb/Uvgv7e//wjw89rIbwLrQoiLx53nVE7lVP5o5AMFBoUQ14BPAy8D57XW9+yq9zHuAhgFcbux2x277B5LRFBjz+15GPQTRqORz4u62TNJEuIoOBRIFI1Zs2kJNLdxP83vtgBJl631Dv5K45rOnj3bAsGEYWi6zsQxyloAcWwZihzDDO18e1VV3k3pdghuzuIGYwBlVSERPj/fDLi5YKYLNrnjHGpZtux5LwDxyCCgW3TTdLF8cNQeVdAOHLZmVXl4dj9kTlsTWix5Ji33rmMJLJrhu7Iwmi8EMqgtoJZF1jivWWfM7yBoYyfyPD9UN5IkCf1+n/l8jhDCu0Orq6veUu0+x8A+79aYOtQt+XDwtvm8giA49Pz8MjtGwYyXo4Kwj60EhBAj4N8Bf19rfdDxl7QQR6ARFh/vx4AfA/OAq6ryL8V9NEkcMrC+lLDswIN+nTax565TUI3ovb0yv76bpnLR0lZDEXvelZU1v20Yhly4cIHt7W3fj8/VOYRBhBLt40opvRJpvqAoiuoX0VA43YHtBn+v10Nr7RVKc9BltspxWafZyh532YfQ/OmvXYg6rdlRmO2IeLN1l6jdGV0fS6jDSsDudOg6WtfQWHcUUKeprD6IEgCgrPdpZobc9S6LE7jsUmJNbJcdMb8bhTKbpQyHK378DIcrXnl3n4VT7u5epJQ+pecuwb3vRR+7UwJ+3On6uFpraLy3JEkWgtqcPFZ2QAgRYRTAv9Ja/3u7+L4z8+3PB3b5XeCJxu5X7LKWaK1/Tmv9Wa31Z8Mw6K4+lVM5lT8keZzsgAD+BfCa1vqnG6t+EfjrwD+xP/9jY/nfFUL8AiYguN9wGxaL1lRFUQe8LEWTo7cC/E9nXjUZcZtmndvG/e3++xm/E/Fu4QuCgPl87s9VFAVaay5dvMzt27fp2VmgG1iDmm3XZTWa5rvLi9e4b/xPba+haw4rZcqlHctRs0txbK/vsItjpKnZl82mTdPXXd8i6c6o3QBe8xjdbeUCK6e7X/e4y6532XG66xYFR7t/++fmO/3U5nczUKtVg1LOGTw2MCiEJAoblpjWPLj/kLObW5yx1Hej4Yg0TUniXmssQD2GWpaAbLskze5OTprbh2Ho4fTu+A5/UTa+p+bYWySP4w58D/BXgW8IIX7XLvsJzMf/b4UQPwq8A/wlu+6XMOnBNzEpwr953AmELVBxDyiKIoTWzGYz/3GNRiO/rpnOcje4yHzUHI4od92DrmkZx7FXMK4H4WAwYDwee9Pctd/yL4KakquqKgr7AlovuBFpdvs4c64bE3CEG+100WFztXndzXtsSnfwNaX5vA7FJTja1O4qjqP27567e95l5+je37Jtms/2OGlud5ziabpg7ud8PvdcgklDGcdxzN7eHs899xz9vpkser0eeZ4zGAwWfoRBEPixAG1FAO2YkntWLSUg6o/b08XbWECWZb5s+bjnc6wS0Fr/Gt0IRS0/uGB7Dfz4ccc9lVM5lZMhJwI2HIYBm2fWKF0e2TbC2N3e9tpwY8OYWMPhsNUow+HamyZP5TX84iDYoqyBsw76/WErquq67F69WnDPkkVkWdZyG6DW2m72bgZtmhYO4KGeUtdavTlrRlHkmXwdaYdzgUKHi8AAp7r3ByBUDS5pZhe62zZdkKbLVGp1aH1T3LG77oCzyLoQY/czYLlpfihQuei+OsfrPt/jpOuGuCDfMmlaYbVFGvlAnVs2m03Jsoz19Q2uXn3KB5CrqmJ9PWQ0GvmMQfOYTUvAzfIti0q2XYeWNSnMstLWvsSROY4LsgsNqW3Qm81TD59fJKew4VM5lY+4nAhLQAqTQpO+8MRo2Z2dHa/5tiz5YhiGDIfD1qzgNKrPZVuL4ijf1s1yDumntSbPc1ZWVtjf3wdgMBghxJRer8fTTz/t/f7pdMrBwQFVVfkqxPv37wONenFo0XRHUUSh2tzvLi3pZv7utYVx5P3LMAw9yaSbTeKw9hlbxT5okAKhDs/mi4JpUkpKrXyAzKX4usGr5nVrcbjAyMdZqrYlsOh3qduWQNcqae7TtToWWQJHxUra91rnz93s7Gbh5v24Z16PD7O83+97avUaJRqgFLz44ks89dTTLSvVNXQdjw2hrauy7J7L3H+drgZ8+rkVwKUT/FbtYi2X5s7z3Hcg2t3dRTVhxR05EUpAaeUj8eDAOAZ0427k3XffBQy//fr6ug8UJkniQT1NmCUsNhfrwWQ/tLBthg4GA/b2TC+CZn52OBx6Zt35fM58Pm9lL9x3UpalH7RuMKRpSpZlHmjjlrsAYjNbURSFJy8pisIrlSZs1QV89vN6EDc/hBb3f9lWPG59N8qvBIeyLK5kuOWuiFp5NQFezWM6xEgz8LboI226B80gmLum7sfeDIo1l3ePvygLUWMPDisTj3kQzetJ/H0CXhm7/otBEPhlFy9epNfrce7cOTY3N7l58yYAm5ubfnLqZqMcbsQ9IzPmy06mqBO4RbXejdCgApsBcP0vy4oiyxmPx+zu7gJGCTTrP7py6g6cyql8xOVEWAKOT8Bpxel0Sn+4wnw+90EWNyM+ePCAhw8f+pnjzJkzbG0ZvnYXZOn1jdmeJAlFUbRm9CAIrBl42Dx20M8zZ86Yy9LaV8q5IJDbZzAYtPZ35lbzXE6c+1J1UIruHN3raDIZHxX8cjOum1EWBjx9I8zatIyiqGVlSClRAh4+fAjAt771LcAEw5y14maSUtVt2br3YUzkii529DOf+QzD4bA12ytl2HAX9QdwLt4iKLF7xt1n1lzWdGMOu0P1+2mi9rrvTMo2ou8QtfyCaw7DkMlkwtmzZw+9hybatCnNQGG3X4NzD4IgIAjb2JSqyAnCgDw32+7vG+t1Npvx4MED8qx+Z48ePWIymRy6bn/epWv+EKVSFQcHB60Bn5eK6XR6KA/qXqyLmpdlyWQyIUkS728NRzWpQxzHng8PjBlmzFxz7iCoAR+LzMowTA5hC9rR78OR8EVgGCEEZdkusXW/dz9eV/Gn5eFjtPa1u3UHsbsPoQ4rBfcMmiW/UkoIajJRxxvojlOW5SFzVtFWAm77phJw7sW1a9eIoqilBLTt8bgo3tD8kP31LZGuEl3200mTkGaRm9jFEbhYwCKF1BwPi0qxnYu6qEKyeazG3XfepQO4VZbAJfckJGmamrFvYw0u5qDLimw2R1WCzGYHyiwnksuzAydCCahKsbe31yqgiOcZm5ubFI0gC5gH16Swckrg4ODAL3dxBEHgMfgelCHNTOg0sPPr3LImAMj97V5m0yeVEoSoZ5pFYJDuC499A5XFA9RJ08dv+fmqzQ/j7qHr7wmPhGsjDd2xHZa9OYhjGdOz3A1uJmumDrsfzCKF2bUO3HUNBgPz3qIaAVcGZUsJdD/MRfUB3RhB81kd9Syb4gK55hzt4b9IqdXb+v5RmJoUcGXUQkAUBd4KbMZChGj3dGinpA/XZZj7Mc8tL4ylVhQFeW56CxaZ7YJVpOhKMZ+bbfJ52rjHCiEC0sz2hNAVg+HydOhpTOBUTuUjLifDElCKWZq3ZoU8zxlPp95vddmAJhc+1OCcoij8bFBH0guKMmtt7/Z3pnDcKBtuzvzu2M4/bUJ4u9u5/aHGdHe3dz6juz+ozcFuSW7TT3QisDx2zbSSI5SkY9ouKPFt+rTu2prpveb1deHAzeMsmoXd+mVRend/zTSWK8fuzuBdoFH3WMvWPb4lUDW2PRzT6N6Pk6JYvm2zVLh5TmehdLd199l0s4zbpVCqXuYsgaoqUGWF0mUrrRkF4SHrxcHVtTbxAXPth9vFN+VEKAFoD36FufBmm6v2CylaqRV3g90B5JRD0xwToq73BtNcFKBoNORsXlM33eiWN9NjzXMuCko1lUn3+MuUQNdHXZSz75aVLvKDu361u/amr9o197uBt8Up1uORek0lI6Vs4Sbc8u6xjjpmd5umy7RMCXSlea9FUTf3WJTGbLpMXYo1FwR167ol1t3rax6zeYxF7lNNAlX3EtRULbdW2p4azvXzbNqVqXfIypz9fZMinM0mR+IETt2BUzmVj7icDEtACGQYErjyYKtli6LwJk1ZOfy9OKRZXVvtykW7+67k1zZm1DU9tJlRoRvQC4LD6SSzg/u9nY4KZM0kBItnp674oFcnk7Bsu+Nm2mUpNBYcv2kJOYRlc11zRgqW1Ay4c7rjdt2BriXSDpAJf13OPfuglkD3+bRAUQtSrYuO1wq26to0XxSIXXRPze2b1kHTbWzev6tEPCoj4SQQAaARHVZQhQuKaz/jSyHQuvKlz+6d5e45aJhPzbdjunAvf64nQwmweMAVpWKemmho/eIEqix9usn5loGoexHObE50av3fKIoIojri70hCoUb6BYFsDOzatCurdtmyu5ZKGf8ss9fnXBIHAV70gasjUFuLnoVcUrzZTNNBbY4e+iDFYXqwVkyh6c7QZlly4j7grhsi9OGYgD+2PKx4Dt3DgtRg9/6W/d2Ux1EazW019ftslvF2lYBTVO4y/a1J0/jGrI/9tddxggrXCCcImnEgfzete2r+DGXUchG0Sz9jP3hRc2cEwl6DPWocWpdYl5QllFXOPJ3Wxw6Xf+qn7sCpnMpHXE6EJeCi283IaWV52xxSrxlZT+fz1gxQVRVBg0zx9m3DczqdTllZWWE0GnkgjAwDkiTxBJJdc7Q7o0pZE390A49uZgFa/IjNbfw9au3dgGWmoT+nj/ovLnv20sAbLIroa6X8FNaM/DsEYPN+mkG7Rbl7jzdA2Cvr1As0LIGqbGMTnDncfCa6qlrWQPfeutH5ZdaA2+6o7EF7h/qci1iS6msWNotknour13CWZDco3MVhLLpOt3/zfpvXUFQOO2H+juJ6W1++7R6JMOPelc27nhhFUSCFIeRxjVHLslyIY3FyIpSAQCJlH2WZfxEBK2ur7E/GvHf/fQA+ZdfNxhPWVlepKvN3Np8TBQFCa2KLitp939Ad/ur/+5/ZXFmDec6q7Q937swmo9HIN2a4cv0aAP3REIRpEuFcsn6/T5SEJIM+eVEgrUuBUOhQMh+PGdjU5dgBM4QmDhLCRvrGRO9DKjug0Eax9fo9ZrMZSdTDvaM8NazGURQhBAShWZHnKVEc0gSToBN//DJL6fWNebq7u0vS7zNLM85fuGCXmcpIn7aUkih0vmuFjJvme+0WuEo4pWqwVlnlxh2rKpJezQbtYcN2wMeJAW+l6YzBYODTvWVZGjctqMlXHVtzmqaeO98hF3d3dxmNRsxn7WKqM2c2TSp5PPEp5KqqCGy8o8nuVFWVzYg0uiM7JVcpAiFtrwibDdCakpJpad3R0CnWkoSAWAhWUrMsrwp0FJCqkkJolJ1ggjBECkEUhAhHA7ZnkH3D/sCnD92zjlONiEOw6NbvvPWeuc9zZxmOViiLjMiO2yiQTKZ7XmFMZgYgFw9jJpMJ48mEuVVcZVkeqRhP3YFTOZWPuJwISwAAoXyk05mZWmufHfB8adaU9SY52nP9O5M3y82sXBQZjx49oKclue0qlFsWGDebfPm3fwOwveQD017MFyL1eojEYN4rNBcumh4qJYL1jTNsXbrA3M5umYWj6tVV8rJoscZ4sJC1VJxOns/npp+Blp7lptfreWx/lqXkDXAIeYnSDTOe2sSsdE2BrRvYg5k1CZtAHQeWynJrmSiNCGSLHcm9h65ZXoNRzIzqz6lr3kH3bprAljzP/XEcRqFZPt7FVLjn45Y1c+RuvesFEcdx61wOG9KEervlULtG7mely0YHI/OzLApKSt+9am7LtmwmT+4AACAASURBVNdGa2SzOSJQzGZm2WhtlXmW0o8iegJy61qVeY5EkBdTSru/K+ia7Oyxt7fnrylJEtLdA95+9zYPLZ/F5edN6frT4nlKVdCLE7S1gLHj1OX/m9iHqqrY29vz38xxlsCJUAJCWPBDo2lFYAeyRz2ldUOPoiiIokaKzg5Gd6Nra4b3/erVq2zfu48sFZVt93zv3j201qyvrwPUvQ7LkiItvSkOMN07YI5ROEEU8tprrwEwnqecv3CBv/A//MV6YPdrwJIu2gAW/9N+S6FPTxZ+ILoBv7+7z2w2wyDISm+GD4Y9pASt6+qwpGcKpcJA2PSRe34lQgTkec6+NT/jXhtQpSsFulEUpPB/l0Ud55BSUpUdZKDUFKowZntRuzxKacIw8Cy+ToEFQUCRt9ubIxRVWd+LVrUyEcL0oJhVZhDLALK0OPQ8J5OZafHVACL1+33TFswCypxCc4p1UQUnUqAr7f1rt65SFaWN+czGJuO0nRYUs5R3375JODZjcjAYsPNom+l4TJkXPqZTliVFmVHadDfAPLMY/0AynU7rVuNSkqUFlQYVmvFx7ulrAAyHffpJD60VhUURaiXoRSF5u9kxRVFQFhXb29utytej+AROhBIAOxM0UkthGNJTtYZvdiAyEGGzn7JBQdGYrS7aGTt9/nl+9da7hAp6loWnPxhwZn3dD4CJnSkJJEJpyrLwxRhKKcqknplc+bAMIq5fv86Vi5forxpfNEiMMklLMwupblBPCu/LqwZzTxAEqFLz/vuGv/Dtt29xcDBGaHPPrjpsbW2VUlUEQZN0wgzwosicDeWfH5ggqLQBIxHUSlMITRxGxFGNYBQyPPSRSSlbTDjup5vFe73eIQScW9+8Dsd710R5SikXWgK1714j8sIwJM9zr5zdNW1v73r0oysac3Tv6+vrrK6u+oDw6uqq6WgVhD61mjrLQEoThFXKjyMpJRLp24NH1nOejiesxD2+/eprTO4ZNqmDnV3ODFdItGQoI3RW2PdSIAJJVpWUdgaYKLMuSmKKLPOI1bIsmZQlSb/HoG8mqI8//zwA589uUWrlq1DBlHn3ovrzDdz7TTOKomB3d/dQM59lchoTOJVT+YjLibAEfIpFNtMoynC8WU3pzOXhcGhTXFZ/WUCQsOYzwAUbEV8djPjOt15j5+EjJjvGz1oZjUjLgtCminLrY+Vz2zykSXkVhlQ2dZRlGfnU9pjb2OS5Z5813ASuTtv5knlOlCQI2eHmCyTa9g5sznDj8YTpeMJ775lI8P7+PkVREoUJGsn+gXGHlBbs7e0hZG32n7EMzOl8hmqgwkxLrBn94YBAmtlzZt0p419XBEL6FtqgfJu3+vnXM7ZjXHbrhBCerKWb6nIzudvfLWvWezhLzx0DanqxNE1ZWVkhyzK/rN/vM5lMGj0eTNbAUX3Fcdzq/5gkCY8ePQLwbt/Vq1e5dOkSayurdctuf/cQCIGQ0scAFNALk5rP0Y7NlXMDIi149sYNvvqeyVzNpzPWZYLOFaqSBNb1TJRGhgGVyn0H435irTgNSkMUGzctHq2QUJKWFc6ic25tlqdkRUEYSc+Lkc4L8qr0RDX+XpRiMpmwt3dAUdTjzLl6i+REKAEnNUlGmz8A8Mwoa2trFt1mzVyH9AJSV3Fo+xfGvYTv/6Ef5LVvfJObb74FGLMum80Zz6b2eMavns1mJnUlpHdBClVRlMb8DOKI+cR8xFevXuW5559nY2ODzJlojeaUWJNYiQb2oND+I3UDTWvN7u4uDx488Gb/6uqI9fVN0nnO3bv3uXXLcCv2+kMePrxP0ov8x9BfMa6ItKZgaFNTcRxTVRVx1Gt1Rqp/BlRF7p9XIA3SrItxcArAxWXcsV1cBtq+pjP7PbW6VRBOgbvjNxmN3DKnwPf29rh8+TLb29ueZj4IgpZSkHLH3G8YMpvNGI1GLQVz/fp1yrJkOp2yvb3tl08mE65fe4oV+85l7JSgQV9KKQis2xgAQSgIHe+ifc+Xzp3n4b37fObTf4zXfvurAFy+dIkVFRCJgl6mGEjzYZdFwSSd048DZjZO4tLESRzT78X18ysLMlWwsrbKpSeeNNcQuomlQKuCSkvivqt6lS2f37k4RVFw79599vf3/TMZDAZHIitP3YFTOZWPuJwQS8ChpOxMiURVCmSNec6suV5Y07Qu3DEtlxwNNMDUpgMHccL3/6kf5Atf+AJTV09wMGY6nrBjZ4h7d0yv1PF4TJHlvs0UmBnxoMrQGA179dkbADz7wgv0Bn2ysvBt0PNZPeOWZenrGPwdKoVyzSRssC7LCp++Ga2aANZwuMLq6jrjg5T40QFCGM2/uXGOMEiI+7FHKV6/bq4nDCV7O9soO1ud3dw0Qccw5O5dc39nz5nO8Xt7eyRRTBDUpc1RGPpOxoBHmg2HQx9U6qb3ptNpCwCklGnVniSJn92dSR7HMYPBwL9LRxvX7/f9e3SW0P3793nhhRf49re/7QO8k8mEXm/gA7OODXptbY379++ztrbmazdeffVVgiBgY2PDp8rMszYNY7a3twmtBTCwgTWFRirDFqSsexfKdiOQYc+yDe8fUBUF6yurXLt2DYDvfPNbpGlKP0yIBFC69wzDJGRcZMSJo4m3WYxhDyGET+Ol8zm94YBzF87z3MdeMO/FWT5hQCRDyjL3zzsMAtI09e/Q0dFnWcG7777LwcFBy0U7DQyeyqmcylJ5nK7ETwA/D5zHTNU/p7X+GSHEPwb+NvDQbvoTWutfsvv8I+BHgQr4e1rrLz7GeVqzTVVVBNS9BNw6BxSqCR1sbj+JkbbFeVoY7VoUGb/76tdYX1ll2DezyHBjlc0LZ7n+gplFe3HdVDIUxs9q5u7FaOCxA84P7g9X2Nvboz8c+Bns4a7xU50/HjSuUQhhqxwtBLhw7atmhGHMxsaGh39KKen3+xQ5jEYrbG6Z2fDZ5z9JlmX0egnzzFg6q2uGFbkqc3a2dzmwjLNxHHNua4sgDDk4MBbQpSumW/zu7i7D/oDNzU0GA5Nqi6OIeZr7WenBAzODb20JVldX6fdrTkcwwcuDgwn9/tA34EjTjDwvieMemU2RvW/h25cvXyaOe40gYGmvs+7WOxiY4/T7E9bWzlBV2h97PJ4iZehnRtfoJUkSnzp0DNErKyueI7GqqlaH6SzLyNPU+9JxoxmMUsYayC2kW0mJENoH3i5eMJbUu2/d5PL5C0x29/ne7/1e87zuvEcwqAgKhdQSi7om0gazMQpXUD3XTm9ox5CxjNx1b22eRa+NOHfhAhcvXwbgzTumBqY3GCFDwcHBga8nGA2Gh1KGYEhFHz16xHQ6b6TXVSt+0JXHcQdK4B9orV8RQqwAXxVC/Ipd98+01v9rc2MhxMeAvwx8HLgE/BchxLNaLw9PKqUYj8c+8unywk30mDO98iJtAYPC0CiM+XzuB1R/YHkJgoBCa7YnB4ytixAEJv7vsgPNwhpHe9VE+qlsbpBnVeHN3Hw6IUhi8rIktjlrF8RyD1toU+7ploVhSGjx8TN7LfNZShAGhKFkaIkg8zwnjGPOX1zn9p0d0tS8yJd/63fZ2NhgNpuwuWkGzrMffwqA8V7OLMvZ3jFMMi9ECYFlWP7yl78MwPnztTuwt7PL+fPnGQxMYLHfi1k/E/PKK68A+J6Lb775Jj/0Qz9Ev99vIc6KouBXfuVX+OQnP8mbb74JmCj8YDDg3LlzHlTlMh6DgVE67v186Utf4tq1a3zuc5/jN37DIDaffvppfxzX6MW5Jc888wzz+dzfw/Xr1wG4efMmL7zwQotabmtri36/TxiaHoAuoKy1Jk1TQ0g7Mh9iPBr494zSxgW1tftRv48QNb7E4RA2NzcRQnDh8iXO2tqGv/63/ib7D7cZRAmBUibsj8lqhL2ESmjWNs340FYZJElCFAV+zEsgCyR7B2MOJmZicfUQeVWSp0ahaXt98yylPxyQWyRiZsfJZDKByuAi3Fh0rtoyOdYd0Frf01q/Yn8fA68Bl4/Y5UeAX9BaZ1rrm5gW5Z8/7jynciqn8kcjHygwKIS4BnwaeBn4HuDvCiH+GvAVjLWwi1EQv9nY7Q5HKw3ApENcSsRBbuMo8iaqQ36tR+scjPfajD7SmNou/5tmjt+/9Igy37fAkoQ4JhbZKQ8WsnZBRFD3ixfC9M8zIn3Jry+3rWrIaRNfDyYtpKuaw74s6v5xYRAiA420eeSAgDCSqAqiuMf9B8bNiOIR+wfvUZQZb98yZuKf/FN/HIBkYJ7NgwfG/C6Kgjjukc7mvofjb/7my4AxpT//+c9x/vx5P9PsPNpGi9oNWl01KbTpdMpsNuPixYsteKtLab7xxhveNN/b2+PjH/84s9nMpxRfeuklwDSIcQhBMNbXV77yFW7fvu1bZbltHz16hNaa+XzOO++8A8CFCxd56aVP+ffy67/+6wC88cYbfPazn+UTn/iED5jt7u6yvr5Ov9/n4ODAW29pmhpUY174MVWzStWcjg7yrEqDznTjIrEW3yBOGMQ9enFMf2QsgZXVVbLr5vxxHNep0LIgzTOKqsAZ7jMLG9ZT0zfAtYnTWpO7QJ51QUo33gJJGARooVCqpj5vVkm6d5ckfZ544gm+/fqbSFFjWGSwfL5/bCUghBgB/w74+1rrAyHEzwI/iYkT/CTwT4G/9QGO92PAj5kbaHftKYvMfChJ4l+u+2jPnzsHQnlTUQQ1TttTholGPbtWNDl6KjSBFp7pxYGDBKAEhtDR185LZx3aWm67k5ZIjJ/vsA1Bo05daw0N6ixp/6eePdYcP0n6BHGADDRhWMNye70+ZRkzGAypbPvyQJlo/5NXr3D/vjHXXT3AbDbhwcNtZpaTfjIxTVR1pfjkJz8JwG//9lcAeOGFF7hx45lW74ak3/MwYHM88xHfunWbGzcecO3aNc6dOweYj+nGjRtsb2/z9a9/3XfWOXfuHM899xwPHjzwH+9l69u+/vrrbG1t8cQTJi5x8eJFbt26xf379/1xm8VLN2/e5Nd//dd9R6T33rvH9evXvSL/+te/7q/961//OufOnfNlx4DvdtTv91ssu1prKuqiJ3e8Jq7BKYaiKCgKxcTiSR49Mgo2iWPiwPQBTBqKXweSvekYGYUENvtQaWVAL0p5rMRKz7ggjhHLjW+tDXuwUUau2MseXErLElVncJQyEX9n5tds3Oa5f+fbb7U4DI5q4PJY2QEhRIRRAP9Ka/3v7UXf11pX2hRg/3Nqk/8u8ERj9yt2WUu01j+ntf6s1vqzjtH3VE7lVP7w5XGyAwL4F8BrWuufbiy/qLW+Z//888A37e+/CPxrIcRPYwKDzwC/deRJbD7WlVs6mGmzBNIhv65cucLq6mod9a0KQ0iapmSl0apuXRAEKDRFVbaompvFLO0edgIlha/11QDSzPYC4QM+0loCgrosuFmy6g22RjFKq5TTIvzifo8wlARhTR4Cxg2JRQ+FJImNqT8e52xurjAabZBbd2JqC53u33/Iu3fveLj0vfffJ89z1tbWeP7Z5wDYsUHDH/iBH2BlZYU8r0uv19bWSNOMJyxS7dVXTS/ClZUh589fII4TT9QaRTFnz25x9eo1vvGNb/Ld3/09ADz55JNcuHCBKIrZ2DDWwcsv/5Z/tqura5yzWIWrV8fcuvUOKysrfOELX2hd3+bmWZv/DnnppU8DcOfOHe7eveufobNY1tbW2NnZodfrtSpHDXOUCbo5JOL6+jpJkjCbTMlsdsatc63TZKUa5rkgjCTSzrQVdpxISV5WlEVBmtqyXikIBz1UFBEM+xR2AOxPxgRBwKg/oLSw7dQVVVW25VivztDE2sDTM4fyjBvEu0qhqXxVac36hH1um/b9JNy69a5HWQK+M/IyeRx34HuAvwp8Qwjxu3bZTwB/RQjxEuZbuQX8HXtxrwoh/i3wLUxm4cePygxAzQlQlLWvrG3/tYFlp3n77bcBUxfQH/Z8jCArDFAoiKIWsaI/dmUAlZWz5LVZ5qi+CrtcSIkIJJU09fiAQXs4JdAw/bXWhoBUa68YXA1BZE2vJtDE7ePKDNzHF0c9wkgiZEkYWTMPbZiWwsikxWILgY4DorjP22+/i1Iu6muO9+6d26yvb/CpTxnT/2uvfJVbt97h05/6FD0LcmnW4rsGqS7155h+HM7emclXr17lypUrrK2teRfBgYHW19d9BB5MHCGOY1544QXv53/pS18C4Lu+67u4ceOGH6gPHz4kSRLOnz/vI/1OySuleP/99318A2BnZ4f19XXvTnz1qwau+9Zbb/nref99g+OfTCaMx2O2trZ8xSkYk9+kWHse4NWEL4cyIFR4l89QdxUULq2rrVcvBKGNISTWhVOY6tNCGNYnp6QrpUBL5vMMZd1Fx+GQi6rFaxH3ElaSxMLNrRYp3ZgwtGMIBZ4Wz8QJHOu1e3dlXvHowUOCwIwf8+7j31uKUGv9azT5tmv5pSP2+Sngp4479qmcyqn80cvJgA1rA/opHQeflWw2J7IzjctHb2xscOnKRa/5tNCIIKDf7/vswnxiLILKuhOmytAcUyhNgEDaSL22mllLYQKDUniOOC0EIrCmvxCeDES7GUPVloCr/CpsQLBJWOrMMRewjJO6MWcYBiA0jhG60iWBjNBaEid9XNjm8pWrDIcrfO2bv8PAFki5AqA333iLZ599lqtPGtzAy1/+Dd5++20+/amX/DW4KL4LULkKRoBez8yYLqDnzOSLFy+yurqKUqrVqTgIAra2toiiyAfvrl275olhnSXwwz/8wwC8+OKLRFHkZyNH/NqM3ju3rCgKvu/7vo8bN27wxhtv2PeoWVlZ8ViMP/En/gQA3/jG1zk4OGB1ddXP+Gmasrm5SeDGRKNpaxiGjPcPPMS2WdgUxKZqU/pmrgotlY+qB2Hs7x8pkUFAnFjyWyDXFSrLPBQcoB/1yOYpB/tjkshYXd95w4xjEdgCIFud2B8O2VgZsba2xorN9pQebGbb6AnDQ1CPnZrwtJkdef/9902g0fXxEOLIwOCJUAJCGjLG5tUMkh6zWeofaLpr0HC3b9+m0qUHgURJRBjHhgpMttN1EgHWjPdkERrChn/vacosAaUSeMSXliaSrzHugieisfsLvZw/vyzLFgKyGXn2RB3CpCM1FYFNEaqiQkuNKkzXWmWV1crKGiKI2No675mSXYYkiiKuXrvO0FYVPv/88ySxAT01uy6D8R1drMV9OGtrq6Rp5kFCL774IgBPPfUUly9f5u7du/6ai6Jgf3/fE3fU5b1Dnz68desWUCsBl35zSmBlZYWtrS0ODg5aVG5gPtZer8fFixe5c+eOX3bz7Xc8WMgBi/r9Hv/hP/wHXnnlFV54weDtZ7MZvV7PVx026xXAuDOu4m5vZsZQHMfoKEY32q0HQYhqUJKldhwqNFoKIhkwz+xAsfdXqorBIGQ6Ma5TJQNee/V1Ht6/z40bBqH6iq08PHN2k9X1Nd/mbjKbI6oSIeuu2K6deBhGRFGAlpo4rtu7S/n/t3c1sZYcV/k71X37vt95M/ZYsZ2YxMa2ZDuJiGWRLELwCrCNMOwiWSILJDYgwYKFUTbZggQLJIQEIkqCENkAIhskfoTFKoYYOY4TZxKTyWRsT2ae38y8N/e9+9ddh0XVqTpV3fcZjzzvXun1J43enb59u09XVZ86/yeWhLt48SIA4Pvfu4AbN25gPInenuNUAaDPHejR49RjJSQBQy5evqp8jbu6wZnNLViLVjDEdDrFwcFBrLO37vL3q6pC6TO17t5yYbWS8VbPG5XDD3DDIRag8pZ6FAbkg0M4xA4QLM9dRqBXIwBf1ssyDKKKoLsFFYN0WJkZtbWQ2A3x7bJ1qorlGKBU1050K6vKWZa9OlQN1mAJ+PgTn8TBLRdAdP6887H/wtO/iIc++lH4x8fTTz+NvXevJZLA/fffD8BJBFK4I/roXVy+nCPGuu3t7RBoJeMvIahnzpzBAw88EDL7xLhIRCFASfcy2N7eDnM5Ho/TeoOAkkp2MJlMsLm5iU9/2gVDXfzRJVy8eBGf+OQT4Rz5K0U0hI69vb1QfFTXFNzY2AhSgojg10cHYX78w4XzJ/MJqHTrAohieM2u7Nm0nmON4jwbY8DzOWAZ0yNfAGfnLjTTGYamxM/c9xEAwCt+7u+991489sTjWFt3ov94NkUzPcRwOERZuHM21qOXqygIKJK2CUluhEhfFy5cwGQyweHhJMy9rsXQBTrOdXBS2Npc5yce/1hSpXZYrQcxEogizXA4DOIkgJbOByARgYuiAIr4kul8BDkHyJp/JMU1bMvSX3pRTKIGgbQ5Sl3XgEmvY4zBxhm3eEWsdTqfcQuNIn3MwK3DMW7ePMBXvvI1AO7FaWoX/77lI9V+7VefAeDi5a1KJqlKR5tt4jHdezFEx6mqQLoAZ3Sxtlucy1i4YJp4fWm/xsxB3NYBQHVdh0g9qS1Y+nqAgPMAyHM+/PDDGAwGwbbw1k9+gt3d3SBSy28effRRvPTSSxiNRnj0YVeZ99q1a3juuedCExD93K4q9TjYAi5fcurG1oZLU16rhoFxSfBZXmglrR0Zx0yKmm5sb2Hdv7yHh4e4fv06jo5GwZ4hHo7h+jDkyMi5R4eHSdpvLKISW+Tp4ik6OvPlbzp37IULF3B4eIiq2ogMzTPVN964+AozP4UMvTrQo8cpx0qoA4y0Tp3spnmHV8DtIk7ccYYeKXeld2rh9JJtyGpXlr955VoJrRRvQjw3bUsmtEh+gzgH8jZetrFh5yMinzXm7imW9sZal+bZ1FBZraiqCuerdexs7eCpT7mYekkNJRRh9zg4cMbSm9f3XN+F0DyTYIwLtdZdkGT8ZIfU2ZI61yGESahdMNTj9/Miv9FWZ2st6roORTt0IJaWNnQXYDnnHe/nl54BOnNxe3sbw+EwzJVIDZcuXXKZdU0Mv5VsRSJqdV4uiiKpkHz2nMuRsHOXatvM66Cq6Sajeu6ZdAuxOnwn59b1DMY4SVQyK+s65mVIPETDFsNhlaTK13PfWFY1mgWcOD+dTkMAnYzB/v4+Nn0mqKS0Hx4egpkSKU3nznRhJZgAkMZtE4ogXuWTYK1Nat6tr69jc3MzqSykLfSAywnIXSRyPWEYerHoc6TEt5QHB2JAkDEm5AxoRjEYVkkt/PzemmEURYH1YhAEy6Jw7sRyMMRg0OCFF14IdBZFgXo2b9UNnM0nTt3wgVXMje+A3G6RHhKpjAFpBVMVrJTrurGsk8pCRK5v4XQ6dS9OWFwmqD1SG1+CcrSnRMYqn9czvnR7VVXY3NxMKjxtbt4dmD4APPaYi4I8vDXCxx9/LFnk58+fx9pa5TeA2GHaGDe2dR3VHXEzj0e6IY0wRgJzbBMWWtQzIGqAVpVc3wf3rLoHwtbWFppmnsw54JPTmMNzSWUrqXsApFZ9UxAMxcKu9byBbTh4dK5e3fW/qf27E3NXqsFa6A/RhZVhAgCcTw5AVQ3SLEG4cEhAQopjPX5ZiM6wGOsRAIpxEDp3c339LiYARFeMDv0tEJtXdunM8jsBEeHWwSF295yOKw1T14YbOHv2LLbO7IRdblBWOHPuLIwpwdaKXQrVwDGec9vnYteaDV/QcjpLcsZNslu1Jz8PYxY7gYx5XacuJS0JyLmhZ4IqYSXSkUAWvhzPm6ZqnV2/OIBjevIMOzs7oYIyEF/ivb09X/RkPUgH586dC4xK19vXu6r8fiybSWFCTQK5h/RPCJmgnI6jY2SpJDC3bifXRU9cVmG8v85gbNiGuRRJRM9N2MR8PwfNnEejI4xGR7h82aXlXL/upMKyrLwx3IYQ42qAFhPS6G0CPXqccqyEJEAggF05JwBBl7MWocSUMXHHdpxeJIUZrD1Iii6KuyhEVCnvgBwLYlVdJ9d114zSgLamaxRIJQFthWbmRF8WfZh8+bNLly6H62xuuS45pZS69vRJJ53g2iFG4YOLRFw8c3Yn0G5AIR1ZaigAcdfU3gvZmWNPvgplGZ9F20lC5FQGeT4tCRhy95UcicbGuvfGUBDNmRlkCEPffRgA5j75yx7VQX8vvBj07t614H0AELpXr28McTQegdGg9MFWe9d3k+KoWiWr6xrzeVSnJCV7MplgNnYqVUg5z3ToOpNMAecu1JD7aSnWuStjAo+u9gOkeS6DwTDZ8U3oBSFqblyvo9ERbtzYx56XLuczd8/NzTVQSbBNLC+m1bkurAQTANoiJzOhaaIuKQ8kYqUW58fjKSaTWRDPxuPYiGIwGMAMyiAqV97/nr/Ui9QB6fGn6SMigEx42fPfidFMvyD+JADRz12Yga9JaHEwcs1R6sbdryjJ18/zU8QNtre3UdezsBCuXCmT6wdDnOHw0rS+QzTWaWNpIjrbaFTS37u5kVDs+FLKnOUqkkRu5rEA+gXoYjzSXFae09VijC+S/B2ur2E0GmF7ezscu3nzJnZ2dlqtzKSHgp6X+TRmExa+QIyMkzQ7zdeFC/+Gpz8eC+HpVLTqYk4msXnLuo8LkI1IZzIaU3p116kMsmadqsahDB/gDINXf7obyooZI2HXtVdBimDzcfUrutJ//Pwu/KZHjx6nAislCQikfZJ214laYC0H1QCIBpymaYKKIHkFoWyUajcu6oCI2Vpczr0A7toA4Cy/FHYAV72ILMN0ZEmL2K+bqRpjsO5dZ7PGeyRQYDAQlxQCzWu+COWZrS3cuLnnr2oxm4ywtraGmc9NbzhKR0VRBAN/M7eoZ2kwj1i3w/NxVInCjulF/7xFue54nAetyD6i54+CRT51y2rpLVex9nZj/4Kjo6Okl8DV3V3cc889gV6xqEvfg8nkKNA8Gt2CtXVipQeiilnXUYSvSonS9HUrvAFO05obfi0QSpDlxtc8CEvKybku2o4+3achVcnKYAyN7ly3to6OjrC/v4/Dw0Mc7Lu1feXKFdy8uR8ik5CHzAAADBZJREFUSqUa0WxWByO6XFvPQRdWggkw4kMAwGw2DlV/i0LCHWOn16ZBKHJRFKm4CkS9aTZzonOtyjIBsRceEH32Ij5LdhYQGYOIudqvnqsU+v6mNslikJdUFgVL8ThqQAUDMCAbXVnTozEOpmOAOESfTaczHDVzjMdHmM2cuLi2se2fc466ji4hSZyxalELmqYJ7iJtJ0kt/VFEdS9savXXdgWZH3nepmkgQQ952a78hTL6s6dTek3qiEQpLS4Vf6UGorbkv7v/brhnnrglNGjmDwDTyTihryxLcCOdozN9f8HzJ+eENRbrAAKxU7KMqUC8GIBTGSQSU2cEAsDB/ggHBwehTqL/Nba2tsKask1kPOOjqStUU7RVvy706kCPHqccK5I7sMGfeOKRRLySQI1FYll6Liciqhbp5vM5asXVJaotjxPQn7UYq/3fWk0Qy32eCluWZTiuc+WLosCa7zsg4ppUg2VOjY5ifDMFhbThwrggIIBjyrSRXc20dmr5Gzrwcpvf55JMeP6Oko/5rtc1djERh1u/6fq97gCdB1AlklSZEpSvWZfj301fMhZaWkEsDmtMRj/acQJ5jENuNHT/2kbEVHVyjVSAKGXpOggH+6MQHdj1nO5+lF3bf6cMvu5DKv0WRYHLb73T5w706NGjjdWwCSjuDDgjGWfGpeiSJUDpqF31/PJowxwLXT+e82u3klyvK+yViELrMx35Jq4xuUewb2T6r/HupLKsEmPc2lrlipgSgiRgCrdjGUNhdy8L30LNpNcl8iWmKc+IdN917coaWn1k5iRWoKVbcn59Ci3YBbkUJ59rLb3Uev5jdSYAQJPeM3cVWnTr53pdBPclCshuKoVp3bw3ya6dR0Bq93XTcLKL6x1fqisBCLH+s1mMHZAw9Tw3oWkaEIqE5jzD1Y9cOOaeyUfHZvs5Kckufz9yrAQTAKIl1X2OATfycA3ii802nVzKXgIdbinndIn/8p2GFt3c4LVTbZkZjaeRlJ9XaMgtzBNp3OEnUDk3Ol+QwNhggy/anePFUH+BYrCePgsUQzJpbMN7GYmITPB+6BRWuW9Oby4SC+Nx1xd6kJyrr5HHXuSeA/l8HBZ9m1vt5fld/EYZGOv+vovNmM8m6iWehWvoOAMpVKtVBKmO3UbWg0LRose+qbN1TDYxTOcMXH9mTjdPoph3Q0Q+dyTScNxY9upAjx6nHCsjCSSRZswAEwjtBB3hzjmXJ6KQghlEI2a1XUQOCtUxIHJZ9gaiPKkoJr8kbsDM7ah9svqZwvMgRtCJEYuIQqSazoAMKcgmF79Tt1XD0ZimMwKDUKViAVppwVk4MKviirkRbPGeG++rd3Vj2qpW+pu2JNB1bldobU6ffN9lLMwNY6KWye8lnqSuZzBFW2LUfvxQDYoZthE3c2zyqQ3U+XPqaEeR1qy1YDtPnjEfAx3t2ZYKGACFJCFAWo41/rjtlCC6sBpMIAsccYvUfe4SY/LBlgfOBy9MDHVfJyUhZgamTCDeMwaMOJVEMyJtP+jKUHQimkyy/M5nhyF9dgmSkbwH/ZwJzVkaL6tr5JbqFlqHKej3VrfXYni1pH09Igqh0MkFKU3pbt3pGHUg3pdb87Dor6ZLB+zoaxSFK9xaZ7YHwMXoE9gVfh3EGAOtDkh4rtzDmBJzlXdijAn9CCJRbi0VJsYr6NDp0hgYnfHqN5qwjn34tmWOthdOm+VInEaq/1uv3qX2g0Xo1YEePU45VkMSQLbDwACUikZxozAoigGI6tbv86IiwQKrmOCieIN854jXJTheyWHTY/biM5sges9UqHNZVC3xjpkxqLy0EPLs3W9dF1ovtvsRaBhAY0FW6CRY5Yt2Dyj0NODkGVU32uBdSQ2buWqhIzZT1SH1KEjVnvA7kybLCD0AoHtMdommiTGUFkgsCY3pbpb8JJ9jaksthALcsOsMlNCn/Pl+WGpfX6DwhWjLUDw27tRVtdF+pg6jr0QjAtFAa6BqYHrUHOsvAEDTRANh69kpSm6ASxyS376XCpdjpZiASMlBfLaxnbeI0LlrRBZcrv/JuY2vsJPf6zgm0CU6JYsY0YJPWctn8RyYxNsRQzqBuPgMmTBhWhxurAWz3NO/kKZASZS8kGJjkOKXYQwUcwr6s1jsGWCk+Q7EadGMhNlx0zFeejy13u4Zjtw76dSMjs/xYJ3p/4mNIVKa0G1FBcjWu0wVc3qM2YIYsEKz8io5D0hMZ8/tDDHnwPcQJJO4COWvQQwq0yXror3EJM+kNyxhAnKOVgv1teI4tis0iTpqbTuYaRF6daBHj1OOlZAEhLfl3NJxurbVeNFOLcfFzx2Cf2y68y3a7fPrOImj3WDUquCh3Poe/OeK8eaWY5FqLFnAxjp48Z6x2xEUnU3jYgcEiRDCsSCGpSgZ5HUOw1kqrJQYodMRgNDOTRsYZcdPn7u7bmORlXcLJGZSlwVaO5SOPwjroUsUzqCbxUba02sb41rKGRYjrsST1GGY47XTgLFShS7LZfV6DRJqZuQF8pqNqRqmVa9hNUyO6UIwIg1EQ2eaxFYNHL1SSJaVZ0gSshZhJZgAI6Z6AqoSjwlNWENPOIEwBym2wMyBm8x9wMe8aWDZelHQv8CwgAGsVLkR8U8aj1gOrkbXbWwexK6wqMh1F9cLdeZ7GroVQoBahGQcaaK6KE0YMADbdrOSdg0I64Nwui3umjkQPO8gzSjawVJhvVP6Omt3ox47/X/322gfEHequlXQvTtfWgSnVjyg6WILZIwwXCdRLTg51OWOlL/OPhQ9T9HTGLP+4vllMr+auRgJxNIeGD9nTIuZH4BWIFBCq29GE9Qov3cZMo5xNUDhbTCGDbjmsI6lCU7lqzTr9Srr9/Jb77TuqZ++R48epxQrIQm8X7yXz79Hj9OOvPTbcViJVGIi2gVwCODdZdOicB49Pcdh1egBVo+mVaPno8x8T35wJZgAABDRt7gj13lZ6Ok5HqtGD7B6NK0aPYvQ2wR69Djl6JlAjx6nHKvEBP5y2QRk6Ok5HqtGD7B6NK0aPZ1YGZtAjx49loNVkgR69OixBCydCRDRrxDRBSJ6k4heXBINPyai7xDRq0T0LX/sLiL6VyL6of977g7T8GUiukZEr6tjnTSQw5/5MXuNiJ48IXq+RERv+3F6lYieVd/9oafnAhH98h2g5wEi+g8i+h4RfZeIfs8fX8oYHUPP0sbotqGz5076H1zk6P8CeAhABeDbAB5fAh0/BnA+O/bHAF70n18E8Ed3mIbPAXgSwOvvRQOAZwH8M1wA7GcAvHxC9HwJwB90nPu4n7shgAf9nBYfMD33AXjSf94G8AN/36WM0TH0LG2MbvffsiWBnwfwJjP/iJlnAL4O4Pkl0yR4HsBX/eevAvj1O3kzZv5PANf/nzQ8D+Br7PBNAGeJ6L4ToGcRngfwdWaeMvNFAG/Cze0HSc8VZv4f//kWgDcAfBhLGqNj6FmEOz5Gt4tlM4EPA7is/v8Wjh/IOwUG8C9E9AoR/bY/9iFmvuI//xTAh5ZA1yIaljluv+vF6y8rFelE6SGijwH4FICXsQJjlNEDrMAYvR8smwmsCj7LzE8CeAbA7xDR5/SX7OS5pbpRVoEGAH8B4GcB/ByAKwD+5KQJIKItAH8P4PeZ+UB/t4wx6qBn6WP0frFsJvA2gAfU/z/ij50omPlt//cagH+EE9Ouivjo/147abqOoWEp48bMV5m5YVcA4a8QxdkToYeIBnAv3N8y8z/4w0sboy56lj1Gt4NlM4H/BvAIET1IRBWAzwP4xkkSQESbRLQtnwH8EoDXPR1f8Kd9AcA/nSRdHoto+AaA3/QW8M8A2Fci8R1DplP/Btw4CT2fJ6IhET0I4BEA//UB35sA/DWAN5j5T9VXSxmjRfQsc4xuG8u2TMJZcX8AZy394hLu/xCc1fbbAL4rNAC4G8C/A/ghgH8DcNcdpuPv4MTHOZy++FuLaICzeP+5H7PvAHjqhOj5G3+/1+AW9X3q/C96ei4AeOYO0PNZOFH/NQCv+n/PLmuMjqFnaWN0u//6iMEePU45lq0O9OjRY8nomUCPHqccPRPo0eOUo2cCPXqccvRMoEePU46eCfToccrRM4EePU45eibQo8cpx/8BNT7RoP9f89IAAAAASUVORK5CYII=\n", 77 | "text/plain": [ 78 | "
" 79 | ] 80 | }, 81 | "metadata": { 82 | "needs_background": "light" 83 | }, 84 | "output_type": "display_data" 85 | } 86 | ], 87 | "source": [ 88 | "# img_path = 'E:/dataset/catDog/catVsdog/test/cats/cat.1500.jpg'\n", 89 | "img_path = './img/car28.jpg'\n", 90 | "\n", 91 | "img = image.load_img(img_path, target_size=(300, 300))\n", 92 | "imshow(img)\n", 93 | "x = image.img_to_array(img)\n", 94 | "x /= 255.\n", 95 | "x = np.expand_dims(x, axis=0)\n", 96 | "predict = model.predict(x)\n", 97 | "# print(predict)" 98 | ] 99 | }, 100 | { 101 | "cell_type": "code", 102 | "execution_count": 5, 103 | "metadata": {}, 104 | "outputs": [ 105 | { 106 | "data": { 107 | "text/plain": [ 108 | "'rana'" 109 | ] 110 | }, 111 | "execution_count": 5, 112 | "metadata": {}, 113 | "output_type": "execute_result" 114 | } 115 | ], 116 | "source": [ 117 | "predict_classes(predict)" 118 | ] 119 | } 120 | ], 121 | "metadata": { 122 | "kernelspec": { 123 | "display_name": "Python 3", 124 | "language": "python", 125 | "name": "python3" 126 | }, 127 | "language_info": { 128 | "codemirror_mode": { 129 | "name": "ipython", 130 | "version": 3 131 | }, 132 | "file_extension": ".py", 133 | "mimetype": "text/x-python", 134 | "name": "python", 135 | "nbconvert_exporter": "python", 136 | "pygments_lexer": "ipython3", 137 | "version": "3.6.5" 138 | } 139 | }, 140 | "nbformat": 4, 141 | "nbformat_minor": 2 142 | } 143 | --------------------------------------------------------------------------------