Skip to content

Instantly share code, notes, and snippets.

@luctrudeau
Created April 7, 2018 02:30
Show Gist options
  • Save luctrudeau/b23c661106f347a576cc9f1959adf23a to your computer and use it in GitHub Desktop.
Save luctrudeau/b23c661106f347a576cc9f1959adf23a to your computer and use it in GitHub Desktop.
Display the source blob
Display the rendered blob
Raw
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# CfL in Action\n",
"\n",
"Although The YCbCr/YUV color spaces are designed to eliminate correlation between color planes. Local correlation between color planes can exist. Chroma from Luma (CfL) exploits this correlation by predicting chroma pixels using reconstructed luma pixel values.\n",
"\n",
"In this notebook, we will demonstrate in a visual way how CfL works. To do so, we will convert an image from RGB to YCbCr and using the luma values show all possible predictions that can be acheived using barrbrain's simple 1D notebook."
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [],
"source": [
"%matplotlib inline\n",
"\n",
"from scipy.ndimage import imread\n",
"import matplotlib.pyplot as plt\n",
"import numpy as np\n",
"import locale\n",
"\n",
"def conv_ycbcr(im):\n",
" y = np.array([0.213, 0.715, 0.072])\n",
" u = np.array([-0.115, -0.385, 0.5])\n",
" v = np.array([0.5, -0.454, -0.046])\n",
" color_matrix = np.array([y, u, v])\n",
" yuv = im.dot(color_matrix.T)\n",
" yuv[:,:,[1,2]] += 128\n",
" \n",
" return yuv\n",
"\n",
"def conv_rgb(im):\n",
" im[:,:,[1,2]] -= 128\n",
" \n",
" r = np.array([1, 0, 1.575])\n",
" g = np.array([1, -0.187, -0.468])\n",
" b = np.array([1, 1.856, 0])\n",
" color_matrix = np.array([r,g,b])\n",
" \n",
" rgb = im.dot(color_matrix.T)\n",
" rgb[rgb < 0] = 0\n",
" rgb[rgb > 255] = 255\n",
" rgb = np.uint8(rgb)\n",
" \n",
" return rgb"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"In AV1, CfL is applied to intra prediction blocks. For this example, we will find a block with highly correlated luma and chroma planes."
]
},
{
"cell_type": "code",
"execution_count": 77,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"1.1102230246251565e-16\n",
"1.1102230246251565e-15\n",
"0.35555796796358186\n",
"0.7872481620530994\n",
"0.8264528681964056\n",
"0.8300872703842255\n",
"0.8346691919774369\n",
"0.8354183192425325\n",
"928\n",
"768\n"
]
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x7f9deda36eb8>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"np.seterr(divide='ignore', invalid='ignore')\n",
"im = plt.imread(\"../images/shuttle.jpg\")\n",
"#im = np.round(im * 255)\n",
"yuv = conv_ycbcr(im)\n",
"height, width, planes = im.shape\n",
"block_size = 32\n",
"height = height // block_size * block_size\n",
"width = width // block_size * block_size\n",
"max_Corr = 0\n",
"for y in range(0, height, block_size):\n",
" for x in range(0, width, block_size):\n",
" block = yuv[y:y+block_size, x:x+block_size, :]\n",
" #avg = np.mean(block[:,:,0])\n",
" dy = np.ones(block_size) * round(np.mean(block[:,:,0])) - block[:,:,0]\n",
" du = np.ones(block_size) * round(np.mean(block[:,:,1])) - block[:,:,1]\n",
" dv = np.ones(block_size) * round(np.mean(block[:,:,2])) - block[:,:,2]\n",
" \n",
" corr = (np.corrcoef(dy.flatten(), du.flatten()) + np.corrcoef(dy.flatten(), dv.flatten()))/2\n",
" corr = abs(corr[0,1])\n",
" if corr > max_Corr and corr < 0.84:\n",
" print(corr)\n",
" max_Corr = corr\n",
" best_x = x\n",
" best_y = y\n",
"\n",
"\n",
"im = im[best_y:best_y+block_size, best_x:best_x+block_size,:]\n",
"yuv = yuv[best_y:best_y+block_size, best_x:best_x+block_size,:]\n",
"print(best_x)\n",
"print(best_y)\n",
"\n",
"\n",
"plt.figure(figsize=(10,4))\n",
"plt.subplot(1,2,1)\n",
"plt.imshow(im)\n",
"plt.title(\"Size: %d x %d\" % (block_size, block_size))\n",
"plt.axis('off');\n",
"\n",
"dy = np.ones(block_size) * round(np.mean(yuv[:,:,0])) - yuv[:,:,0]\n",
"du = np.ones(block_size) * round(np.mean(yuv[:,:,1])) - yuv[:,:,1]\n",
"dv = np.ones(block_size) * round(np.mean(yuv[:,:,2])) - yuv[:,:,2]\n",
"\n",
"plt.subplot(1,2,2)\n",
"plt.scatter(dy, du, label='Cb', color=\"b\")\n",
"plt.scatter(dy, dv, label='Cr', color=\"r\")\n",
"plt.legend(['Cb', 'Cr'])\n",
"plt.title(\"Correlation between Luma and Chroma\")\n",
"plt.xlabel(\"$\\Delta$ Luma\")\n",
"plt.ylabel(\"$\\Delta$ Chroma\");"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The prediction built by CfL is based on the reconstructed Luma pixels. In this example, we will use the original Luma pixels as the reconstructed Luma pixel\n",
"\n",
"In the AV1 version of CfL, the Chroma block is first predicted using the average pixel values of the neighboring pixels above and to the left of the block. For this example, we predict the chromatic planes using the average value of each plane."
]
},
{
"cell_type": "code",
"execution_count": 78,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x7f9ded22db38>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"pred_u = yuv[:,:,1].mean()\n",
"pred_v = yuv[:,:,2].mean()\n",
"\n",
"luma = yuv[:,:,0]\n",
"u = np.ones((block_size, block_size)) * pred_u\n",
"v = np.ones((block_size, block_size)) * pred_v\n",
"\n",
"def error(im1, im2):\n",
" diff = im1 - im2\n",
" return np.round(np.sum(np.multiply(diff,diff)))\n",
"\n",
"pred_rgb = conv_rgb(np.dstack([luma, u, v]))\n",
"\n",
"plt.figure(figsize=(5,4))\n",
"plt.imshow(pred_rgb)\n",
"#plt.title(\"Prediction\\n(Luma + Average Chroma Value)\")\n",
"plt.axis('off');"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Next, we show all possible predictions CfL can make, using barrbrain's simple 1D code book. For each prediction, we measure the squared error and pick the best one."
]
},
{
"cell_type": "code",
"execution_count": 79,
"metadata": {
"scrolled": true
},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x7f9dedffbf60>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"mags = [-.5, -.25, -.125, 0, .125, .25, .5]\n",
"M = len(mags)\n",
"\n",
"ycbcr = conv_ycbcr(im)\n",
"\n",
"zm_luma = yuv[:,:,0] - yuv[:,:,0].mean()\n",
"\n",
"plt.figure(figsize=(20,15))\n",
"\n",
"min_Err = 1000000\n",
"\n",
"for x, alpha_cb in enumerate(mags):\n",
" for y, alpha_cr in enumerate(mags):\n",
" alpha_cr = -alpha_cr\n",
" u = pred_u + zm_luma * alpha_cb\n",
" v = pred_v + zm_luma * alpha_cr\n",
" \n",
" rgb = np.uint8(conv_rgb(np.dstack([luma, u, v])))\n",
" plt.subplot(M, M, y * M + x + 1)\n",
"\n",
"\n",
" \n",
" err = error(im, rgb)\n",
" if err < min_Err:\n",
" min_Err = err\n",
" best_a_cb = alpha_cb\n",
" best_a_cr = alpha_cr\n",
" \n",
" plt.imshow(np.uint8(rgb))\n",
" #plt.title(\"Squared Error: %d\" % err)\n",
" plt.axis('off')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Since a correlation exists between the chromatic planes and the luma plane, we can see that CfL produces a considerably better prediction."
]
},
{
"cell_type": "code",
"execution_count": 80,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x7f9dee168128>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"plt.figure(figsize=(15,4))\n",
"plt.subplot(1,3,1)\n",
"plt.imshow(im)\n",
"plt.title(\"Original Block\")\n",
"plt.axis('off');\n",
"\n",
"plt.subplot(1,3,2)\n",
"plt.imshow(pred_rgb)\n",
"plt.title(\"Prediction\\n Average Chroma\\nSquared Error: %d\" % error(im, pred_rgb))\n",
"plt.axis('off');\n",
"\n",
"plt.subplot(1,3,3)\n",
"u = pred_u + zm_luma * best_a_cb\n",
"v = pred_v + zm_luma * best_a_cr\n",
"rgb = conv_rgb(np.dstack([luma, u, v]))\n",
"plt.imshow(np.uint8(rgb))\n",
"plt.title(\"CfL\\na_cb = %0.2f a_cr = %0.2f\\nSquared Error: %d\" % (best_a_cb, best_a_cr,min_Err))\n",
"plt.axis('off');"
]
},
{
"cell_type": "code",
"execution_count": 81,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAkEAAAEYCAYAAABbbSIPAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4xLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvAOZPmwAAIABJREFUeJzsnXmYVMXV/z9nNoYB3ABXZBENGoxBQYUoxi3uxiVqRIyKCwqviSQxiyGJxoTkNa/6i0kEHRMRdaIxUaMmbiGKGndxQVxQiaioEMFE2ZeZ+v1Rt5k7zb13unt6nfl+nuc+3bfqVtWput19T1edOseccwghhBBCdDWqSi2AEEIIIUQpkBIkhBBCiC6JlCAhhBBCdEmkBAkhhBCiSyIlSAghhBBdEilBQgghhOiSSAmqAMxsrJk9WGo5hBCFw8wWmNnBOZYdbWbzCiDTQDNzZlaT77pFcTCzS8zs5g6Uz/lzWQlICSoTzGxfM3vCzD4xs4/N7HEz2xPAOdfknDukwO2fbGbzgvb/bWYzzGyTIK+bmf3ezN4xs2Vm9qKZHZ6ndg8ws5fN7L9mttTM7jSz7UL5l5vZm0G7r5vZafloV4h0zOwUM3vOzJab2Ydmdp+Z7VtquaIIFJMdU+fOucecc0NKKVM6laBAmdn+Zraw1HKUEjPbxMx+ZWbvBp/9+cF5n1LLVgykBJUBgbLxV+A3wBbAdsBPgDVFFONxYB/n3KbADkAN8LMgrwZ4D/gisCnwQ+A2MxuYh3ZfBQ51zm0GbAu8CUwL5a8Ajg7aPR24ysy+kId2hdiAmX0L+BXwc2AroD8wFTgmh7o2euiXsyIgui5mVgf8AxgKHAZsAowClgJ75VBf5X3OnXM6SnwAI4D/JuSfAfwzeP9dYHnoWAfcEORtCvwe+BB4H6/EVOcgT0/gRuDehGvmAF+JyZsG3B46vwz/RbN22u0G/AJ4NeGau4Fvl/qe6eg8R/C9WQ6cmHBNN7yS9EFw/AroFuTtDywEvgcsAm6KSguuPQp4Efgv8ASwW6iNBcDBwfu9gCeD6z4EfgvUBXmPAg7/B2E58NVUe6G6dgFmBeVfAb4cyrsBuBr4G7AMeBoYHNPvgUFb44N+fwhcGMqvAr4PzMc/OG8Dtgjy3g3Kpn6rRgHvAMOD/LFB/tDg/CzgL+3VG+SPDMbvv8BLwP6hvFnAT/F/7JYBDwJ9YvrXZtzS8mYBZ4fOzyD4HQ7OHTAR/8dtWdDm4ECuTwOZU/dsc/wf3Y+A/wTv+yV83lJ9X4b/o3hcuhzA5UFdbwOHh/IHAY8EZf8efHZujmnnbGAx0DNBlgXAhfjf/E+APwL1cZ/9IP0c4C3gY/xv9rbFGLecvv+l/gHS4cBr30uBGcDhwOZp+W2+fKH07fE/TIcH53cC1wI9gC2BZ4Bzg7z+wQ9G/wQ59g0+5Kkf2ENirtsKWA3sHJPfALwRyD0aWNLOFz4lWwteqTsj5rru+B/hw0p9z3R0ngP/D3g9UJNwzaXAU8H3qm/wg/3TIG//oPxleGWpe0za7sC/gb2BavzM5gJalakFtCpBw/EP+hq8IvIaMCkkjwN2DJ3vT/AwB2qDB9APgDrgwOBhMyTIv4HWf/o1QBNwa0y/BwZt3RL8rnwO/0BKyXlBMC79gn5eC9ySVrYmVN+NBH9igEb8g35CKO+bGdS7XSD/EXhl6UvBed8gf1ZQ72eCcZ8F/G9M/zaMW0TeLNpXgu7C/34Pxc/c/wM/k74pXnk5Pbi2N/AV/G9jL+BPBApfTNsn4mfGq/BK7gpgm5Ac6/CKRjUwAf8csCD/SeDKYNz2C+59nBJ0KzCjne/HAvyzZFv8SsVrwHkJn/0D8b/5ewRpvwEeLca45fT9L/UPkI4NH4xd8D9OC4MP1d3AVkFemy9fkNYdmA18LzjfKvgwdQ9dMwZ4OAdZtgMuAT4TkVcLzASubaeOvfH/At4BxmTY7hb4fxQjY/JnAPfTzoySDh3ZHPgZiUXtXDMfOCJ0fiiwIHi/P7CW4N9xQto0AsUplDYP+GLwfgGBchHR/iTgztB5khI0Gv+vvCqUfwtwSfD+BuB3obwjgNdj2h0YtLVzKO2XwO+D968BB4XytsE/oGuIVoLOAu4OlT2bQAELfiv2yKDe7xHMOITyH6D1wTkL+GEobyJwf0z/NoxbRN4s2leC9gmdb/g9Ds6vAH4VU/cw4D9ZfEZfBI4JyfFWKK8hkGVr/B/K9UCPUP4fiFeC/k6Mghi6ZgFwatr9vybhc/574Jeh857BvRtY7HHL5JBNUJngnHvNOXeGc64fsCte6/5VQpHfA/Occ5cF5wPwCsqHgZHxf/H/nrbMQZb38crGreF0M6vCT/WvBc5vp46ngX8Bhp/ezKTdj/GKzl3pa8tm9n/4cTnJBd8GIfLEUqBPO/YM2+If0ineCdJSfOScW51WJj1tAPDt1Pcz+I5un1YPAGb2GTP7q5ktMrNP8bZKmRqqbgu855xrSZN3u9D5otD7lfgHVRLvpdWVknkAcGeoP68Bzfg/ZVE8Aow2s23wsxi3AfsE9oWb4h/27dU7ADgxbRz3xStKufYvVxaH3q+KOO8JYGYNZnZtsLnkU/yS5mZmVh1VqZmdFmxASfVvV9re/w39c86tDN72xN+X/zjnVoSuDX9u01lK23GLI2k80z/nbb4rzrnlQTvhz19Bxi0XpASVIc651/H/1naNyjez7+Ones8KJb+Hnwnq45zbLDg2cc4NzVGMGvw6bapNwyteW+FtgdYlFTaz/8FPhX6At2PKpt0t8VOlqbp+gl8mPMQ592kWdQmRCU/ivzvHJlzzAf7hm6J/kJYiSjFPT3sPmBL6fm7mnGtwzt0SUXYa8Dqwk3NuE/zSlrXTj7Cs2wd/WsLyvp9h+Si2T6sr1ff38Mvx4T7VB3+kNhoT59xb+Ifo1/FLJJ/iH7Dj8bMsLRnU+x5+Jiic18M5978d6F8UK/CzLCm27kBd3waGAHsH93O/IH2je2pmA4Dr8H80ezu/aWRu1LURfAhsbmY9Qmn9E66fCRyadn22pN/nNt+VoO7e5Pb5y3jcckVKUBlgZjub2bfNrF9wvj1+KeupiGsPB76BN5RblUp3zn2INwC8ItjyWGVmg83sixnKMNbM+gfvBwBT8Ou0Kabhl+yODrcbU9dn8EbZpwJfA75rZsNirj3ezIYE8vbFr2W/EMwKYWYXAafglwmWZtIXIbLBOfcJ8GPgajM7Nvj3WWtmh5vZL4PLbgF+aGZ9g63DPway9b1yHXCeme1tnh5mdqSZ9Yq4thfeSHS5me2Mt/sIsxhvQxHF03hF47tBP/bH77C8Neb6TPhRMC5DgXF441iAa4ApwW8GwfikdtR9hLfzS5fzEfwD/pHgfFbaeXv13gwcbWaHmlm1mdUHW9375dq5oI7wYfhZqeODfu9I2z+d2dILP8PxXzPbArg44doeeMXio0C2ccT8IU7HOfcO8BzwEzOrM+/i4eiEIjfhlcrbg+dQlZn1NrMfmNkRmbQZwS3AODMbZmbd8LOYTzvnFuRQVzbjlhNSgsqDZXgbmqfNbAVe+ZmL14LT+SreMPM18z4dlpvZNUHeaXhDyFfxlvR/JpjqNLP+wbVx/wo+CzwRtP843lbhnKDsAOBc/HrsolC7Y9MrCZYUbgYuc8695Jx7E/8v9qbgC5HOdvilt2XAy/gfzeNC+T/H/5N5K9TuD2L6IEROOOeuAL6Fd//wEf7BcD7wl+CSn+EfLnPwn9PnaXUhkWkbz+G/U7/Ffz/fwtt3RHEhXvlfhlee/piWfwkwI1guOSmtnbX4B9/heAPVqcBpwQxzrjwSyPsP4HLnXMp561V4+8UHzWwZ/rdr70COlfg/U48Hco4M1dULv7QRdd5eve/hXRf8gNZ79R1yf55th3/Qho/BwP/DL/0vxi/TN+VYP3jThu74+/EU/jcvEufcq3i7mCeDtj+H/03OlFNotcm8GG9wHtfWGuBg/Kzj3/GK9zP4pbens2gzXOdM4EfA7fiZqcHAybnURRbjlispa3IhhBBCiC6FZoKEEEII0SWREiSEEEKILomUICGEEEJ0SaQECSGEEKJLUtRgZ0ceOCrWCnvZ8nj3L83NzZHpa9em+yZrvwzA4B0HxeaNHj06Nm+nnQZHpi9evCgyHcDvtIzJq4o3Sq9KUE+rEjLNoutMKpOkC7uWJJ9UCX2L6XfieBQxr7o6vl9JmwVyzWtpaYnNiyuXa1tfPW1S3nxodAb69OnjBg4cWGoxhBBFZPbs2Uucc33bu67yIr4KIUQWDBw4kOeee67UYgghioiZJXnK3oCWw4QQQgjRJZESJIQQQoguiZQgIYQQQnRJpAQJIYQQoksiJUgIIYQQXZKi7g5b+vGS2LyWlvWxeevXR+ctXxG/rb579+6xeZtttklsXkNDfWze8uXLItN79uwZW8YRv1U/mey3VOdOvC5cU1ubUK78t8jHkcuW9Y7kJZFbOcX8qxSammDyZHj3XejfH6ZMgbFj49OFEMVDW+SFEKJANDXB+PGwcqU/f+cdf/744zBjxsbpIEVIiGKi5TAhhCgQkye3KjopVq6Exsbo9MmTiyebEEJKkBBCFIx3341Oj3NoH3e9EKIwSAkSQogC0b9/dHpc1Ja464UQhUFKkBBCFIgpU6ChoW1aQ4O3/4lKnzKleLIJIaQECSFEwRg71tv/DBgAZv61sRGmTo1Ol1G0EMWlqLvDqix+u3VtXV1snnOrItPXr4tvq7ZX/Bb5LTbfMjavpjp+i/yiRR9Fpvfp0ye2jHO5bcVuaYnfWt8cZ1BA/Jb8XLeEd+vWLTYviXLZIp9LFPkkirl9Pv+uEEQpGDs2WrmJSxdCFA/NBAkhhBCiSyIlSAghCkRTEwwcCFVV/nXiRP9qltnRp4+vQwhRGOQsUQghCkCUo8Rp07KrY+lSGDfOv9fSmRD5RzNBQghRAKIcJebCunVyoihEoZASJIQQBSCfjg/lRFGIwlDU5bA1qxOCiVr87qX166PL1XfrEVumb5+tY/O22nLb2LxNN9kiNm/t2uhArkm71JKCjCZt/nEJ5SBhZ1NMnZao7+YarDW+XLyMuZTJvZxz0Z+dpB12yeORdF+SZIwf/zgZ9R+lsunf3y+B5asuIUT+0a+sEEIUgChHiblQWysnikIUCilBQghRAKIcJU6Y4F8zpXdvmD5dRtFCFArtDhNCVBxmtgBYBjQD651zI0orUTRyiChEeSMlSAhRqRzgnFtSaiGEEJWLlsOEEEII0SWREiSEqEQc8KCZzTaz8aUWRghRmRR1OWyLLXrH5n366aexeRYjZq+e8cE9e/bYLDbvk09WxOa9//7i2Lxlyz6JTO/dO36Ltln8FvPEwJ8J6qklZJpFy1KVIAdV8Vu7VyxfniBIfDmL2bZuVQnjkbDVPddycTJ2r493r0DSfUlsK2H7vIu/Z1XVMXkJZZLGvouwr3PufTPbEvi7mb3unHs0lRkoRuMB+pf5/vKmJrjgAu8dGrwx9EknwU03QerrZwbnneejzwsh8odmgoQQFYdz7v3g9d/AncBeafmNzrkRzrkRffv2LYWIGdHUBGee2aoAgX8/bVqrAgTep9i0aT72mBAif0gJEkJUFGbWw8x6pd4DhwBzSytVbkyeDGvXZn59Y2PhZBGiK6LdYUKISmMr4M5gObkG+INz7v7SipQb2YbDSHR0LoTIGilBQoiKwjn3L+DzhW5n1Sro3r2wbWQbWqM6KbKMECJrtBwmhBBpfPghDBkCV12VHOOvo0yZAnV1mV8/XvvghMgrUoKEECKNhgbYYw+YNAnGjGlrpJxPxo6F66/3O8JS9O7tw2v07NmaZubTtDtMiPxS1OWwVSvjw63/9z/xvzJxEb/ru8dbFH7ySXx9//rX27F5DQkRD5ubo6PIb7rpprFlkrZ2V1UlbJtOUE9zKZfrdvwdBidtL07YIh+zVT/ZZUCurgayjz6/9dbbZl2mvbaSZEz6vxFXLpd+ifyw6aZwxx3wy1964+WXX4bbb4edd85/W3GhNaTwCFF4NBMkhBARVFXB978PDz4I//437LmnV4QyoakJBg70Mzj5OiZObK23qsq/NjW1bS89XQiRjJQgIYRI4KCD4PnnYehQOOEE+M53YH30pDDgFZDx47MzeM6EadPgtNN8vc751/HjvXKUai+cLkVIiPaREiSEEO2w/fbwyCNe4bj8cvjSl2BxjHP5yZNh5crCyNGStgq6cqX3HZTe3sqVXg4hRDJSgoQQIgO6dYOrr4Ybb4Snn/aG0088sfF12fr+6ShxvoOKLYcQlYiUICGEyIKvfQ2efNL7EPriF+E3v2m7jb7YocrifAeVecg0IcoCOUsUQog06upgXfxm1jZ84xv+KBVxM0FHHFFcOYSoRIqqBK1YsSohN35SqmfP6Ijf9fXxXsaWLVuWkBcfRT59zT1Mt261kelJW+4htyjyuUYJT96mnT2vvzYvvq2kIOcx9zOXMh0pR8x41HdLcAWctB0/RxmrkrbWx7k1SOxXfPyECZMujS8n2iUbBaic+f3vYZ99orffCyE8Wg4TQogQnUEBAh+YVcbRQiQjJUgIITopMo4WIhkpQUII0UmRcbQQyUgJEkKIELXRpn9lRSbR5OvqfIBWIUQ8UoKEECLE2rXlrQgNGAAzZviAqnF7K3r39oFZZRQtRDJSgoQQIo21a73vn44eH38MRx7p6zz1VFixouN1LljglZupU/1u1qhrliyRAiREJhR1i3zStvWkyOi1MX/L4qLLA3z66aexeatWx2+RX7s2PjJ90pb8OFzYi1oaLQn78VtcfHCipHJx7SXJ4Vz8ODY0NMTmJREfGT0hmn1ipPjcysWxds3yHNvKLS/p8x1XLrGtKkWRrwQ23xzuvtsvS118Mbz0ko9Ov+OOpZZMCAGaCRJCiIJSVQU/+hHcdx+8/z6MGOEVI0V+F6L0yGO0EEIUgUMPhdmzfST6Y46BmprWaPSpyO+gZSwhiolmgoQQIoGOzNikypr5Y9AgrwhBqwKUYuVKbzeUujZ1VFW1Pa+pgT59fHrPnn6nWCp94sQ8dVqILoJmgoQQIoamJj9Ds3KlP89mxia9bK6km/M1N8PSpf79ihVt06dN8++nTu1Ym0J0FTQTJIQQMUyevLESs3JlZuEoosoWg8bG4rcpRKVS1Jmgmpp4naumJl6UNWuiA68m7Q6rr6+Pzauti/c0lrSLJ2mHVRyJQVITqK7KVcboMUmWPb6+XNoKciNTLSHKaG1t/GegW7dusXlJ9zruczV48ODYMtUJnujidiq2J0cu8ie1VZ3wXRL5Iy7sRCbhKEoVsiLhZ1EIkYZ+SYUQIoa4sBOZhKMoVciKTLxJCyE8UoKEECKGKVMg3VVWQ0Nm4SiiyhaDlM2SEKJ9ZBgthBAxnHrqxmmpXVxRecWguho228x7o25ogFWrvOfo6mqvAMkoWojM0UyQEEJEYAbNGC15OtLrSj//hAa6d/e7wVauhDPP9HIccgh89FFrSIz1631YjJYWWL7c2wCl0qUACZEdUoKEECKCZgyDoh29WMWiVQ2Y+Rme66/3cjz4IPTtC9tsAwcf3OoXyAy6dWv1GSSv00Jkj5bDhBAigpRyks/62jvvRfROWIBFi/wRZu3aVp9B8jotRPYUVQnqVh+/5TeJ9evXRaavWbsmtkzctvr2SNqmXVsbvc153fp4OXINCpq0TbuqKilQZ7T8iQE8E+YDBw0aEJvXvXv32LzNNtssMn3zzTePLZOU17Nnz9i8XLbIJwXzzfW+JLl5yCUvqS2ryt5dg+j8pHwYSQkSIjO0HCaEEGVE2Fbo3/RhDNmtcZXKP5EQlYiWw4QQIgLDu/zM55JYJm2G2+vLUq5nHAC3kNn0Tqn8EwlRiWgmSAgh0qmrA4qrAMVRzzqaOJXmYEfZOmp4kx34HHMwWvgxl7Ceqg2zR2+/Yz6yqqykhWgXKUFCiIrDzA4zs3lm9paZfT+vldfVwbpoO8RSYfgf6yqghmZ25G2eYiSncSOXcglH8Tf+w+atM0krVrSGpFeoeSFikRIkhKgozKwauBo4HPgsMMbMPpu3BspMAYqjgVVMZxzXcC4PcSDDmc3z7L7xhStWeKdC0BpqXoqQEIBsgoQQlcdewFvOuX8BmNmtwDHAqyWVqgQYcC6NjOEWXmEoy+nJh2zNNixKLjhtmj/Azw7tuCNstRUsXgxvvw1r1ngnRIMG+fRsSKojnJeipsYraSlFrarKH+vX5y5DocnHOJUL2falmH0fNgx+9avC1B1QVCXIufUJefFbfh0x/8ws/h9bTW38an5SdO4ePeK3fcdtCe/Va+vYMnXdktrqEZu36aa9YvPitp8DbLLJJpHpPXvFBzGqq4v/GKxbtzY2r7om+zFO2iqeHLE+/vOxvnlFbN7a1S2R6ZtsFj/2uZO0bT3+s9oSk9eSFA28a0cK3w54L3S+ENg7fIGZjQfGA/TvApbCm7CMETzHq3yWeQzhUzZhJ96kKvEzGbB+Pbz+OrzxRqtbavAPuXnzfHpKQQkrTFEsXtz2+jVr/HmKefNa6w+3HyasEIXL5/qgzfdDO6mPlaYIZduXztT3AM0ECSE6Hc65RqARYMSIEdk5VaqtrZglsTC1rGdX5vJjLuXnTGY4z/FnTmAg72RWQUvEH4awUgReYZk/H374w2hnRAMHblxPS4sPcLZ8+cYKUKZyrVoFs2ZlX7apyXuQTM08rVnjfQgccADce69/37+/j3bbnnOlpibvhOmdiPFsaYE334wfl3DZ6mq/LDlgQGbttldP6rWqqnXse/b0Ct/HH/v+HXEE3HijXxoFbx9WV9d2Ri7cl9df92OeLl/S/T3nHC9bNmNaBsgmSAhRabwPbB867xek5Ye1a70iVIFU08IUfshdfJm32JHhzOYBDslvI+vW+YddU5N/KIZjdsQ5KXr33VbX1rkQV29YhnQD8IMPhtNP9x4kw6xcCddc4xUJ51pdbUftpkvVbwZf+1q0ApSiuRnGjds4jklKEUuVbQ6mcZPajWLixLYypOpJvYaVk+XL/Xin+jdtWqsCBD49SgEKEyVf3H1IXZvJmJYZlrTMkG9223GH2MaS5Ghuif5XtnZt/FJNc3P8ekH+l8Pil1a0HNaWQiyHJd3rlqh/tySPfaUzdtz3ymFnd8EwsxrgDeAgvPLzLHCKc+6VqOtHjBjhnnvuudwaGzgw+cHnBcpthqPAvMVgjucO5rIrP+FiJjMls+WxTGloaKtgNDRA9+7Ryk5qtiJXBgyABQvapqWUi3QlJ19t5KP+pDFJUV0NM2Ykz5o0NXkFqBSfs969vYLZ3vegvTquuqqoM0NmNts5N6K96zQTJISoKJw3LjwfeAB4DbgtTgHqMFOm+AdZEv37+wdZmbEj83mKkYyliR/zU77M3fyH+D9QWRM1wwLR49URBaihwd+HdCZPzp8CBBvPcuSj/pUr258Ba2727gz69ImfOZk8uXSK9tKlHVOAUnW018cSISVICFFxOOfudc59xjk32DkX8YTME2PHQmOjnyUAP+sTJvWATkUuLTMaWMWNnMbVTORBDmE4s3mBYYVrcOlSP175Ugp79vT1pc8gNDV1/MGcTroBfbHjj6QUhbCjy4kT/dJevvtaKpYuzWyZLGqptUBICRJCiCTGjvXLJM7BTTd5hcjMv6Ye0FOnwoQJGytJaTiS9xAWAgMmMo1H2Y+11PEFnuAGTm+9oHfvVqeKHW4s6H/MMnTWRNWTWqbKlerqDR7BNxA121SqXYVhR5fTpnVsFq0cWbnS9y/OV1XYhqoI9kVFtQn64l6fi20syVYkzr6kuiZeh6uOiaYOUN89Oho8JEcrb2iIjla+2+d3jS2TFEU+yQYmOVJ89hHmk6PIx8u4evXy2LyWBJcHcbY4SZ+3JBmT7LiS8pIisceR6K6hAHm5jBUufqxOP+fiTm0TlC0dsgnqCHH2RL17+900+VzKyZB/05cx3MJDHMR4ruXX/a+g2zvBFudM7J8yITVrlkldDQ0wahT84x/J9YVtdfIhZ8rOJWknU5RNUMr+q3dvb1i8PP73UWTAhAn+D0SYuPsbZReWgGyChBCilMQtp3z8cesSW2pGacKEjc/zabzfrRvU1rIlH/EAh/J9fkEj5zK65olWMfO1/PPuu5nZUoEfh5kzfX+T6ks6z4WPP/YP1JYW/xplsBteCk3dl5tu8krQkiVeERIdo7Fx47SkHYYFQEqQEEIUgrjllP79W5fYUg/hqVM3Pk/51Uk6UrMucZjBzTfD6tUwfToMGECNtfCLAddy56RHmLekD3vsAX//e4K8ufQ73ZYqigEDWpWPqVPjr02XK07OAQPaH4/26kgn/T6FlaVi2wx1RqKW+pK+NwVASpAQQhSCqNmQuJ1O+WwjhRmcd17rgzvtgX7s//sizz4L22wDhx4KU/b6Cy3d02afams3tp+pq2udBYkzFA+3d/PNmY1DpuOVdF0mM1D5ugddwBN5SSjG9yaElCAhhCgEwWzI8t4DaMFYwADGrmzETh27wQ65Vy+vR+RydO8OPc8dyykrG1nAAFqA9VTTAixgAKe4m6hpnIqZN7OYOLHV719NjX895BCYNAnGjIEf/mkYx+78Ov/d/nMbln/+efZ0vtHr+qB+Y3nvAXD99X45KMlQPGIcinJdVF76UmNUnbkQp3ClbBsHDPAK4M03R8sD7RrSt6G6Gg46KL/LpMUg22XDTD8HeUKG0WnIMLotMowufJ4MowtLyQyjyb8/v0LQ0ADXXgv//S9885v+mXP77TB37sayNzQU9HlUeaTCWOQaKiKqPGReZ1MTnHZa/nbj5ZOqKh+qY+zYvBk7Z0OmhtFSgtKQEtQWKUGFz5MSVFhKqQTla8NVoUk9i554Ak48Ef7zH6/wRPn5K+BzS+RClKZdW+tnURKiKlBXB2ed5WOohWOR5cMDelgBipOxwBp1pkpQUQOoDt1159i8hh7RCgbEKybdE5SZmpq62Lzq6vhnRF36+nebOqMfqGbxWniSgmFVSeVis0jQgYjTI3JVgjbfIjoMBySHq1gXE4AyqUyuClJ1dW7KZLwcSbn51y9y+yMiPacSqBRCkwn2AAAgAElEQVTb2ZScX/gCPP88nHxyfLzSSulTlyGlRLQ3o3TEEZkFjY0K0pqiuhr23x9efLFVQ+7RwytOKXcBUSEy4mQsgylF2QQJIUSGJDmynTjRp4ftdsowpFgkzrX2Z6ut/G6xmDCEbLFF7u00NfnICanx6dmzNd5onz4bxx7NB0V0Plw6onaxtbcDMU4BCTsHXb++7W7E9eu9S4OUTZhzXvlZtqz1fMmSeJcDmbRfZIo6EySEEJVK+ox+ypEtwOOPe+e+lUy4Pykn2Gec4Z97YT791I9Fts+wpiY488y2KzQrVrQGNw8vvaXLkitJ96xMnsGixBTVJmjimSfGNlbZy2Hx9SUtNSWs8CTmVVfH28DkezmsW7f48Sif5bAkG6nsJzuL+Z3Ivb34z9zJX/uB1spC5MsmKMm2c+HCzhPdIGzz06dP/uyCcrGP6qj9UQnscUWZUJY2QUKIzoeZbQ7sBGz4J+Oce7R0EhWGJEe2lbLslQnhfn78cfQ1uRh752JL1FH7oyI7HxYViGyChBA5Y2ZnA48CDwA/CV4vKaVMhSLJkW2+gqaXA+F+xvW5pgZefjn3egtZJpPy8nMoUkgJEkJ0hAuAPYF3nHMHALsD/y2tSIUhyZFtR4KaJxEV8Lw96uq8T71cSHfMG9Xnbt38hqC9987OyHjKlOz6kg8nwUV2PiwqkKIuhw0ctF1sXkOCq/Nem0R7yEyy30mys1ifbukXork52pYliSQbIxL8BCU5C020M7J444O4co74PuPi2/roo09i83Lx7ZRUJhf7HYj3swPxtklJ45sryfcs+/aSy5SN2c9q59xqM8PMujnnXjezIaUWqhAk7fJN5V1zTe5LY3V13r1LylA4tdM43OYWW/gwYKlr0neghXcnT5y4sbF2XZ33Uv3xx5ntmo7r84EHwle/CqeeCk89BVdc0b6Ck6rrggva7q6ur/fypHadpWTLxw7qMt6ZLcqEohpG//Knk+INozutEpTgCyhnJSiXvCQ54utbtmxFbF4lKEFxeZ1ZCfrqqRcVTUMyszuBccAk4EDgP0Ctc+6IYsnQHqV0ltiZWbcOvv99uPJKGDUK/vQn2C7+f64QRSVTw2gthwkhcsY5d5xz7r/OuUuAHwG/B44trVSiGNTW+hmg226DOXNgjz3g4YdLLZUQ2SElSAiRF5xzjzjn7nbOJfjq79pEOe5rz5lfU5N3KphyMFhd7Ze6kupNBUtNOSEMl08/+vRpbTMXx4InngjPPuuXsw4+2AdjHTAguo5U/eEgrunyhR0olqtzw87kgDHbvnSmvoO2yAshOoCZjQAmAwMI/Z4453YrmVBlSpTjvjPP9DY9KdO1dGd+UfExW1pabX2mTo2uN2wLFOXnJ8zSpV6Oxx+HGTNycyy4yy7wzDNeCbr11tb0cB3QVs6UX6V0+cIOFMvRuWFncsCYbV86U99TyCYoDdkEtUU2QZnTRW2C5gHfAV4m9CFzzpVN2NBysQnKxllgyplfUpnqau/NOV9BWtPDRKXLkgkDBkT74BkwwL/mKmc5OTfsTA4Ys+1LJfW9LJ0l9ujZPTavW7d4UaqroxWalpb4WfckRSduxxCAI/udV9U18R6ck5SPpJ1jSRqSJXl/jimXFM0+aVW0vj7ek3d1dfw9i1N2conqDsmKjnNJedHpSf0qBHn/s5EQRb7IfOScu7vUQlQC2TjoS12bVCalsOTL8V+cx+ts6n/vvY7XUYjy+aQzOWDMti+dqe8pyuaXVAhRkVxsZr8zszFmdnzqKLVQ5Ug2DvpS1yaVSf2fyJfjv7j/J7nInU7PnrD99tnLlIsMhaYzOWDMti+dqe8ppAQJITrCOGAYcBhwdHAcVVKJypQox30p30Bhws78pkyJjweYssWIqjdb6up8fR11LBglS02NDzJeX++PbCk354adyQFjtn3pTH1PISVICNER9nTOjXDOne6cGxccZ5ZaqHJk7FhobPT2E2b+9frrYfr0tmmNja1GpmPHwo03eqeCKaqqYMIEbxQdV++ECa3nvXu3LZ9O795ejqlTN64nLEuufbzhBrjlFnj/fe9tequt/LWpmad0+Xr08Gm5ylBoovpYbjJmSrZ96Ux9T1FUw+ir/99FsY116xZvV1NfH20AnWQ4WkyboLq63GyCkqK3J/UtMSJ8nm2C1q6Jl79cbIKS7nVcue7d4+3TCkExbYJOOOXCYhpGTwf+zzn3arHazJZyMYzu6rzyCnzlK/DWW3DZZfCtbyVvDhGiI8hZohCiGIwEXjSzeWY2x8xeNrM5pRZKlB9Dh/pt9MccAxdeCCed5JfJhCgl8hMkhOgIh5VaAFE5bLIJ/PnPcPnlPuTG3Llwxx3ez5AQpaCoSlC3bvFLIbW18Xnxy0YJy0kJS011Cdvxa2q6xebF+b5ZtSrel07yslZufmWSlsPi8pJc8JjFj323uvjxaF4fP8bNMftt165JCOSaQFKf62rjrS3jyiVvuY/vV1JeLv6KkupMlqM81hGcc++Y2eeB0UHSY865l0opkyhvzOA734E99/RBWPfc09sknXRSqSUTXREthwkhcsbMLgCagC2D42Yz+3pppSp/ksJcpM6TQl2kjrjQGwcf7A2Pw6EosglvkElohFzCJ4TLnHEG/PCHsNtuXhmqqmoNp5EeFiST+nr2bO1zTY0fg46GdyinkBIdGe9wmJLU5yN1ns14d0aKahj9u6k/jG0sbpYFoKY2e10tbiaiPWpqkmar8jwTlGAjXC4zQVWW35mgXO9LUp9z8ULdmWeCvnbW5GIaRs8BRjnnVgTnPYAnyylsRrkZRqeHHugodXVtQ2/EUV3tw2K0t5MnSr6Ghra7gDK5JtN69947OvBqeAdcpvUl0Z58mcobV0cuY5Ip+RrvJNob70ojU8NoKUHpbUkJaltOSlDGeV1UCXoZv01+dXBeDzzrnPtcsWRoj3JTgvIV5iIXMglvkElohFzCJ2Tb71RYkGzlTCKb8A7lFFKiHMa70ijLsBlCiE7HdOBpM7szOD8W+H2hGjOzS4BzgI+CpB845+4tVHuFoJQhBjJpO5PQCLmET8i2383NfoYr7v9gLuOYS+iSQqVnQ7HGuysimyAhRE6Yn678E95r9MfBMc4596sCN/3/nHPDgqOiFCAobYiBTNrOJDRCLuETcun3mDGwfHn+6stHCJBShJQoxnjn6MKt4pESJITICefX6+51zj3vnPt1cLxQarnKnXyEuQgTFXojiurqzMIbZBIaIZfwCXFlDjoo+vqRI+FPf4K99oLXX8+sviTyEQKkVCEl8jXeSaTCsHQ1iroc1qtXr/hMS4i2ngNJdiK50tISbaPRo0dCvyqBBLOw1WuTrC2TdOgYz9UJXqaTSLJcW9ec8NmJyUv0XJ3kxjYhqyrByKsT/9t43sz2dM49W8Q2zzez04DngG875/5TxLY7TMqQdfJkv2TRvz8ccQTce2/b8xtvhBXx5oaAtwlJPQjD9e24ozc2Tpmi9egB116bmYFulHxTprQtm8k12dQ7caI38m1u9sra+PHeSPehh+Dkk/02+htu8B6n4+praIBVq3yfq6th//29d+pM5ctlHDpyfTZ0dLzfecePSXOz/8zsuCPMmrXxeHdFimoY/cebLktw7JNfJaiYJBlTVzpr1yXdl8p9tOcavqMSOOmUbxfTMPp1YCdgAbACrya6juwOM7OZwNYRWZOBp4AleL34p8A2UbHKzGw8MB6gf//+w98plSWy6DALF8IJJ8DTT3tP07/4hd/WLUQSCpshhCgGhwI7AAfSGkH+6I5U6Jw72Dm3a8Rxl3NusXOu2TnXAlwH7BVTR2MQ2HVE3759OyJOyYjyCxP2+xL2A9Snz8a+horhF6cY9OsHjzzi+3f55d7HzeLFHauzoz6O+vTxR7r/nbh64vz1ZDu+UfWkXsOfh169vHxxfqeqqqC+vq3PqbCPpVzkK8fPTiZoJigPaCao8tBMUH4ws27AV4CBhJbXnXOXFqi9bZxzHwbvvwns7Zw7OalMuW2Rz4QoHy+1tf7htHZtZnUU2i9OKbjpJjj3XNh8c28v9IUvZF9HIX3uRNWTVDab8c23f6lMyFS+cvzslKWfIClBlYeUoMqjyErQ/cAnwGxgwyZb59wVBWrvJmAYfjlsAXBuSimKoxKVoHz5EiqkX5xSMWcOHH+8l/WKK+DrX88uGn2hfe6k19Ne2UzHt1T+pfLlW6rYyE+QEKIY9HPOFS2IqnPua8Vqq5Tky5dQIf3ilIrddoPnnoPTT4cLLoCnnoLrrvPG35lQaJ876de2VzbTukt1H/LlW6pcqdy/8kKIcuAJMysb79CdhXz5EiqkX5xSstlmcOed8POfwx//6ENvvPFGZmUL7XMn/dr2ymZad6nuQ758S5UrRZ0Jqq1LWoLovMsTlUxnXTbqrP0qFkG4DIf/DRlnZv8C1pCH3WHCb3/Oh01Qtn5xouw68uHnphBUVcFFF/nt82PGwIgRPjbacccll8uln1FlooiqJ6lsNuObqQz5JFP5Ku2zE0YzQUKIXEjtAjsc2BE4hDztDhPemLSx0dtUmPnX6dPh+uv9e2gbD7B3bx8AM3x9tkapUW2Wk1F0HAcfDLNnw847e1uh738/OQZWLv1ML9O7tz+g1dNyXD3hsplcn4kM4XpSr+HPQ8+eXr5U/yZMaLtcaAbd0sJCpsrnIl+lfnagyIbRd/zx8uI1JvLC+uai2dgWlc48E/SVr04q+E0zsx2BrZxzj6el7wMscs7NL7QMmVKJhtEie9asgUmT4Jpr4IAD4NZbYcstSy2VKBXyEySEKCS/Aj6NSP80yBMUx3dKvnzQhOuK8k0U1YdM+1eM67p1g2nT/LLMrFmw1VawzTb58Z/UHkn3IF3WdHkmTmz1OZTy+ZSSceLEtv6ADj64rb+f8JHuNyiqrVS9YZni6os7amp8vZ0FzQSJRDQTVHkUaSboWefcnjF5LzvnysZYulQzQcXwnZIvHzRxdUXZIaXqhcz6l+k45OO6KJnSKeY9yNaOK1xuv/3gH//Ij4wpGhr8rroZMzpuWzRhQnmH2ihLP0FSgioPKUGVR5GUoDedczvF5L3lnNux0DJkSqmUoGL4TsmXD5pM6kqvFzLrX6bjkI/r4mRqr86OUCr/PbmSiiGWj3qSbK9KjfwECSEKyXNmdo5z7rpwopmdjXec2OUphu+UfPmgyee1mfrJKfR1SRTzHpQb+VCA8llPqSnuFvna2mI2VzSaO8unIYLaqs6pJ1dVyRyug0wC7jSzsbQqPSOAOqCdTcpdg/79o2cI8uk7Ja6NXNpqr66oejPpX6bjkK/rMulDMe9BuZHPmaDOgJ4EQoisCQKZfgH4CT58xQLgJ865Uc65RaWUrVyYMsXbYITJt++UqDZybSuqrtpaqKuLrjfT/hXzuqTxCLPbbvmbyUhqM2r8MqG2Fg46qGNyRdHQ4O2XMhmj9hg/vuN1lANSgoQQOeOce9g595vgeKjU8pQTxfCdki8fNHHyhn0Tpfch0/4V87qovLD/pP79/fb5e+6Bww6Djz7KbrzbGzdoew+ixi/dn9OECa0+h8C/nz4dZs70eWF/QAcdFB8eJN1vUFRbjY3emDk8RpmGGwm3U+5G0dlQVMPoe+64qlMaRnfm5TDXSc3GOvNy2DFfOb9zWrPniPwEiXSmT/cP8i23hD//Gfbaq9QSiXwjP0FCCNHFKKZfokK2UWjGjYMnnvCzGvvu6x0s5nM+IDxGffr4oyN+ljIl3a9QlD+fpqa2folyPbp3j06vrm7bbrl/Xjrn33whREkws32BMc65/ym1LF2NdH8177zTardRKJ84hWijWOyxhw+3ceqpflboySe9s8WO2sukj9HSpa154fGC/I7lxIle/hTNza3nqaWrpiY488zs/RZFsXp1dHpLS2u7++xT/p8XLYflgWKOYbGxqhys+iqAzrwcduQx5xV1OczMdgdOAU4CFgE7O+c2LaYMSXSV5bBS+iXKZxvFpqUFfvpT+MlPvMH07bfD4MG515eJ36Bs/CxlSk1NtLF32J9PMX0aVVdDv36l+7xoOUwIUTDM7DNmdrGZzQOuA5YA+zvn9gY+Lq10XZNS+iWqNF85Yaqq4OKL4W9/8/0YPhz++tfc68tkLN59N/9jGWeaGk4v5n1qbq6Mz4uUICFELrwOHAGc4Jwb4Zy7zDn3dpDXeadGy5g43zf59olT6DZKxeGH++WxwYPh6KPhRz/KbRt9JmPRv3/+xzLOb084vZj3qbq6Mj4vUoKEELlwPPA28KCZ3WRmR5tZ5/SGWiGUyi9RvtsoJYMGweOPw1lnwc9+BkccAUuWZFdHe76KsvWzlClxfnvC6VOm5Oa3KFd5KuHzIiVICJE1zrm/OOdOBnYE7gPGAwvNbDqwSUmF66IU2y9RodooNfX18LvfwXXXwSOP+OWxZ5/NvHz6GPXu7Y9c/SxlytSpG/sVSvfnM3as91sU9kuUK/X10elVVa3tVsLnRYbReUCG0ZWHDKPzj5ltDpwInOycO7AUMkTRVQyjRf557jk44QT48EP47W/h7LP9w1yUPzKMFkIUFefcf5xzjeWkAIlWiu2vpdz9w2TCiBHeTuiAA/zyzllnwapVudWVqe+gQpPtfUldn6n/oJoaX3c4rU+f8r3/RfUT1NLSUszmikZ1Z4kkF0F1jVxJCVHpFNu/T2fyJ9S7t985duml/njxRe9leocdMq8jU99BhR6bbO9L+vWZEGVMvnSp908U104pKepy2F1//n+dct2ocytBMQu/omw5/OjxmrAPoeWw4vv36Yz+hMArQ6ee6mc3br7ZG05nQqa+gwo9Ntnel3z7FSrm/ddymBBCCKD4/loqwT9MLhx5pF8eGzAAjjrK+xfKZBt9pr6DCk229yXfMpXj/ZcSJIQQnZxi+2upBP8wubLDDj7u2Gmn+eWxo45qu7wVRaa+gwpNtvcl3zKV4/2XEiSEEJ2cYvtrqQT/MB2he3cfif6aa+Chh/w2+tmz46/P1HdQocn2vrQndzbU1ZXn/ZcSJIQQnZxi+2upBP8wHcUMzj0X/vlPH39sn33g97+PvjZT30GFJtv7Er4+U6qrN3Yj0Lu3909UjvdfhtF5QIbRopyQYXRbZBgtCs2SJXDKKfD3v/tt9L/9bbwzQVEcMjWM1v7nPNCZHe9ZJ/UM1pkdXAohikufPnDffd5QesqU1m30AweWWjLRHp336S2EEEIUiepqH2/s7rvhrbe8ndD995daKtEeUoKEEEKIPHH00T7cRr9+3o/QpZd6myFRnkgJEkIIIfLIjjvCk096x4oXX+wVo//8p9RSiSikBAkhhBA5kBSHq6EBZszw0dT//ne/PPbCC5mXz0WGTGKSJbXZ1OTLR8X8Ss9LP3r2hIkTKy9enAyjhRBlhZmdCFwC7ALs5Zx7LpR3EXAW0Ax8wzn3QEmEFF2eTOJwmcGECbD77nDiifCFL8C0aXDGGfmJr5ZtTLKkNgHGjYN161rPUzG/Hn/cb/9fuzZelhUrfN+S2i9HtEU+D9TW1pZahIJRVd2t1CIUhM68O6zSt8ib2S5AC3AtcGFKCTKzzwK3AHsB2wIzgc845xIDF2iLvCgE2cbh+ugjOPlk71xx/HhvNB0VRiKb+FrZxiRLkhni66quziw8SHvtF5Oy3CLfWZWFzrxFvrPSmZWgSsc59xpEumc4BrjVObcGeNvM3sIrRE9m28a6detYuHAhq1ev7qi4nZr6+nr69evXaX+7O0K28bb69oUHHoAf/Qj+93+zrzfXa8PX5BojLFcFKJO6S42Ww4QQlcJ2wFOh84VB2kaY2XhgPED/iIBFCxcupFevXgwcOLDT+sLqKM45li5dysKFCxk0aFCpxSk7+vePnjlJio9VUwO/+AXsvTccfzxE/RfbYgs/Y/Puu76uKVPil5O22KL9uGXOtXqpbmjwy1ZxMhdiJihV98SJ3vt0c7Ovb/x4by9VajSFIYQoOmY208zmRhzH5KN+51yjc26Ec25E3759N8pfvXo1vXv3lgKUgJnRu3dvzZbF0JH4aMceC//3fxuHl6irg08/9cqIc612NXEGzsuWZS7v0qXRChD4rfxTpkDUhF9dnZehJocpk9R4TJzo7YVSilRzsz+fODH7OvONlCAhRNFxzh3snNs14rgrodj7wPah835BWk5IAWofjVE8HY2P9u1vw3XXtSpS3btDjx5tDZPBGzFPnrxx+cmTow2Vc7HOuPdeL/f06X7GKEUq5tfUqbDppsl19OjhjcCjxqOxMbpMXHox0XKYEKJSuBv4g5ldiTeM3gl4prQiia7M2LEd2/l01ll+99XVV8M3vwmrVkVfF2VXE2drk1r+ysbsMVVXUn8+/ji5juXL4/PiltI6YmuULzQTJIQoK8zsODNbCIwC/mZmDwA4514BbgNeBe4H/qe9nWHlTHV1NcOGDWPXXXflxBNPZGVq33IOzJo1i6OOOioy75lnnmG//fZjyJAh7L777px99tmsXLmSSy65hMsvvzznNkV+MIPzz4dHH/W2MlFE2RnF2R71759sl5Rp/blcE0dcv8oh9riUICFEWeGcu9M518851805t5Vz7tBQ3hTn3GDn3BDn3H2llLOjdO/enRdffJG5c+dSV1fHNddc0ybfOUdLB+MtLF68mBNPPJHLLruMefPm8cILL3DYYYexLBtjElEURo2CX/964+WsODujJJukKVO8LU8m1NVlZseUdM1nP5tcNuyLKJP0YqIt8nmgoz9U5UxLJ91Kri3yIsWkST7qdz4ZNgx+9avMrx89ejRz5sxhwYIFHHrooey9997Mnj2be++9l3nz5nHxxRezZs0aBg8ezPTp0+nZsyf3338/kyZNoqGhgX333Tey3quvvprTTz+dUaNGbUg74YQTNrx/6aWXGDVqFEuWLOG73/0u55xzTs59Fh1n4kTvefnrX/cG0nV1fjdZ1BJVKm3y5PidZBdc0Lp7rHdvOOkkuO22tmlXXZXZkl7qmtNPb7uM9dnPwiuvJJdN7QLT7jAhhBBtWL9+Pffddx+f+9znAHjzzTeZOHEir7zyCj169OBnP/sZM2fO5Pnnn2fEiBFceeWVrF69mnPOOYd77rmH2bNns2jRosi6586dy/Dhw2PbnjNnDg899BBPPvkkl156KR988EFB+igy57TT4JNP4PbboVs3+OlP4R//iL527FjviLClxb+GlZmxY2HJEm8b5Jx/P3XqxmnZ2DSNHQvr17eWd659BSjF1KmtZdevLw8FCGQYLYTo4mQzY5NPVq1axbBhwwA/E3TWWWfxwQcfMGDAAEaOHAnAU089xauvvso+++wDwNq1axk1ahSvv/46gwYNYqeddgLg1FNPpTGHrTbHHHMM3bt3p3v37hxwwAE888wzHHvssXnqoegIxx8PQ4f610MOgZ/9DL73vdx2f8XR1ORnksL+gaqq/Hb41M6zqiqvZA0Y0DrTlCr37rt+CW7VKn9NVZXf5bZiRatvoXC5bOXKxFdSR5ESJIQQJSBlE5ROjx49Nrx3zvGlL32JW265pc01UeWiGDp0KLNnz+aYY6LdL6VvgdeW+PJiyBB4+mk45xz4wQ/8+xkz2t+ungnpccRStLS03XqfsvZI+Sx6/HEvQ6pc2PdQS0vreWrJLNsYYvmIqZYNWg4TQogyZeTIkTz++OO89dZbAKxYsYI33niDnXfemQULFjB//nyAjZSkFOeffz4zZszg6aef3pB2xx13sHjxYgDuuusuVq9ezdKlS5k1axZ77rlngXsksqVnT/jDH7ztzt/+BiNGwMsvd7zeyZM3VoDaY+VKb9eTS7koX0eZypVN+WyREiSEEGVK3759ueGGGxgzZgy77bbbhqWw+vp6GhsbOfLII9ljjz3YcsstI8tvtdVW3HrrrVx44YUMGTKEXXbZhQceeIBevXoBsNtuu3HAAQcwcuRIfvSjH7HtttsWs3siQ8zgG9+AWbP8TMvee0d7kc6GXGN65erbJ9P2co1vlitFjSL/4N+u6ZRbcjrzTqMWVwaOHERWVHoU+XwTFUX+tddeY5dddimRRJWFxqq8WLQIvvpV71fo/PPhiisy3w4fJpMI9FHkGkcs02jySZHus4lGn2kUec0ECSGEEBXC1lvDzJk+7MZvfwv77w8LF2ZfT5SfofZoaPD2ObmUy8QXUZxc2ZTPFilBQgghRAVRWwuXXw5/+pO3D9pjD3j44ezqCMc+C1NV1XZmKbUbLRULbOrUtjHTevRovaaqyp9DqzfobGOqdTQmW7Zod5gQQghRgZxwgt9G/5WvwMEHe8eK3/nOxtHp48g19llHY6aVuv4wmgkSQgghKpRddvFb57/yFe9H6IQTvLdpkRlSgoQQQogKplcv+OMf4cor4a67YM89M/fk3NWREiSEEEJUOGbwzW/CQw/5maC99oIY91EihJQgIYQoAdXV1QwbNoxdd92VE088kZXZeqALMWvWLI466qjIvGeeeYb99tuPIUOGsPvuu3P22Wd3qC1R3uy3Hzz/vDeWPuUUH0Q17AFatEVKkBBClIBU2Iy5c+dSV1fHNddc0ybfOUdLKmZBjixevJgTTzyRyy67jHnz5vHCCy9w2GGHsWzZsjbXrV+/vkPtiNLS1OT961RV+deHHoKzz/bLZL/+NWyyiY/+Hr4mztliqi4zH0PMrPWoqfEG2H36tKb17OnbSZ336RNdd7qMHXX2mC+0O0wI0bWZNAkyjMWVMcOGZRWZdfTo0cyZM4cFCxZw6KGHsvfeezN79mzuvfde5s2bx8UXX8yaNWsYPHgw06dPp2fPntx///1MmjSJhoYG9t1338h6r776ak4//XRGjRq1Ie2EE04A4JJLLmH+/Pn861//on///rGhN0R5ExVra9w4r5CkZoDWrIEbb2wtExePK72udKeIzc0bR7QPxw4DWJJPT/wAABBXSURBVLrUtx+uu9jxwLJBM0FCCFFC1q9fz3333cfnPvc5AN58800mTpzIK6+8Qo8ePfjZz37GzJkzef755xkxYgRXXnklq1ev5pxzzuGee+5h9uzZLFq0KLLuuXPnMnz48Ni2X331VWbOnCkFqIKJirW1bl37S2BR8bhyiScWxbp1besudjywbNBMkBCia5PFjE0+WbVqFcOGDQP8TNBZZ53FBx98wIABAxg5ciQATz31FK+++ir77LMPAGvXrt0QP2zQoEHstNNOAJx66qk0NjZmLcOXv/xlunfvnqceiVLQkZha6WXzGZ8rXFex44Flg5QgIYQoASmboHR6pFzu4u2CvvSlL200UxNVLoqhQ4cye/ZsjjnmmMj8cFuiMunfP7cYYADbbJO/utLp37/9esPXlAothwkhRJkycuRIHn/8cd566y0AVqxYwRtvvMHOO+/MggULmD9/PkDsctb555/PjBkzePrppzek3XHHHSxevLjwwouiEBVrq7Y2s6CqS5fCbbcl15ULtbVtY30VOx5YNkgJEkKIMqVv377ccMMNjBkzht12223DUlh9fT2NjY0ceeSR7LHHHmy55ZaR5bfaaituvfVWLrzwQoYMGcIuu+zCAw88QK9evYrcE1EoomJtTZ8O11/fNm3ChLbnv/41DB/uI9J/61vejic9nlgq/leK6mo46CDo3bs1rUcPv0MsRe/evv2wwXOx44FlgznnitbYg3+7pniNFZFijmGxaXHV7V8kyorDjx6fYeSgrsGIESPcc8891ybttddeY5dddimRRJWFxqrzsnatjzX261/D6NHe63T6ElmlYmaznXMj2rtOM0FCCCFEJyfKT8+f/uTDbAA89hhsuy1svXX7Pnyamtr6CurVy5+Xmw+gTJBhtBBCCNGJifLTc+aZ4JxfBguzeLH38+McnHpqdF3jxrUtt3y5P1J1l4sPoEzQTJAQQgjRiYny07N27cYKUIp16+Dcc1sVm/S64sqlKBcfQJkgJUgIIYToxOTij2flSh+E9fXXc6urHHwAZYKUICGEEKITk4s/ni23hCVLYM894fbbs6+rHHwAZYKUICGEEKITE+Wnp67O+/OJoqEBrrzSR6PfdVc44QS/i2z9el9XXLlw+XLwAZQJUoKEEKJELFq0iJNPPpnBgwczfPhwjjjiCBobGznqqKNKLZroRET56bn+eu/PJ90nUNiHT79+8Mgj8D//A5df7iPIH3SQLxf2FdSzpz8vNx9AmaDdYUII0Q5NTd7Q8913/TT/lCkd/5F3znHcccdx+umnc+uttwLw0ksvcffdd+dBYiHaMnZs9Ge2vc9xXR389rcwcqTf9bXHHn5r/ZIlhZGz2GgmSAghEkhtL37nHb9tOLUFuKO+UB5++GFqa2s577zzNqR9/vOfZ/To0Xz66acceeSRDBkyhPPOO4+WlpYO9kKIjnHqqfDUU95D9P77eweLncFPsJQgIYRIIGp7cT62AM+dO5fhw4dH5j3zzDP85je/4dVXX2X+/PnccccdHWtMiDyw227w7LNw5JFwwQV+FilqG30lISVICFFWmNmJZvaKmbWY2YhQ+kAzW2VmLwbHNcWQJ26rbyG3AO+1117ssMMOVFdXM2bMGP75z38WrjEhsmCzzeCOO+DnP/dhNkaOhDfeKLVUuSMlSAhRbswFjgcejcib75wbFhznReTnnbitvh3dAjx06FBmz54dmWdmiedClJKqKrjoInjgAe9hesQIuPPOUkuVG1KChBBlhXPuNefcvFLLkSJqe3E+tgAfeOCBrFmzhsbGxg1pc+bM4bHHHuOZZ57h7bffpqWlhT/+8Y/su+++HWtMiAJw8MF+G/0uu8Dxx8P3vue30VcSUoKEEJXEIDN7wcweMbPRxWgwantxPrYAmxl33nknM2fOZPDgwQwdOpSLLrqIrbfemj333JPzzz+fXXbZhUGDBnHcccflpzNC5Jntt4dHH4UJE+CXv4RDDoF//7vUUmWOtsgLIYqOmc0Eto7Imuycuyum2IdAf+fcUjMbDvzFzIY65z6NqH88MB6gfx5c18ZtL+4o2267LbfddttG6eecc07+GxOiQHTrBlOnevugc89t3UY/alSpJWsfzQQJIYqOc+5g59yuEUecAoRzbo1zbmnwfjYwH/hMzLWNzrkRzrkRffv2LUwnhBBtOO00v42+Wzf44he9f6Fy30YvJUgIURGYWV8zqw7e7wDsBPyrtFIJIcJ8/vPw3HNw6KHw9a/D174GK1aUWqp4pAQJIcoKMzvOzBYCo4C/mdkDQdZ+wBwzexH4M3Cec+7jUskphIhm883hrrvgpz+FP/zBL4u9+WappYpGSpAQoqxwzt3pnOvnnOvmnNvKOXdokH67c25osD1+D+fcPaWWVQgRTVUV/PCHcP/98MEHfhv9XbGL3aVDSpAQQgghCsIhh8Ds2fCZz8Cxx8IPfgDNzaWWqhUpQUIIIYQoGAMGwGOP+Zh7v/iFtxf66KNSS+WREiSEECVi0aJFnHzyyQwePJjhw4dzxBFH8EYlxyAQIob6erj2Wrj+evjnP/02+qefLrVUUoKEEKJ9mppg4EBv6DBwYMdDyAPOOY477jj2339/5s+fz+zZs/nFL37B4sWLN1yzvtLc7wrRDuPGwRNPQE0NjB4N11xT2m30UoKEECKJpiY/j//OO/7X+p13/HkHFaGHH36Y2tpazjuvNQTa5z//eZqbmxk9ejRf/vKX+exnP9tR6YUoO/bYw9sJfelL3tP0GWfAypWlkUVKkBBCJDF58sa/0CtX+vQOMHfuXIYPHx6Z9/zzz3PVVVdpaUx0WrbYAu65B37yE7jpJr+Nfv784sshJUgIIZJ4993s0vPAXnvtxaBBgwpWvxDlQFUV/PjHcO+98N57MHw4/PWvRZahuM0JIUSFERd7rIMxyYYOHcrs2bMj83r06NGhuoWoJA47zEejHzwYjj7a+xcq1jZ6KUFCCJHElCnQ0NA2raHBp3eAAw88kDVr1tDY2Lghbc6cOTz22GMdqleISmTgQHj8cTjrLP/VOvxwWLKk8O1KCRJCiCTGjoXGRu/sxMy/NjZ2OKy8mXHnnXcyc+ZMBg8ezNChQ7nooovYeuut8yS4EJVFfT387ndw3XXw6KN+eey11wrbZk1hqxdCiE7A2LEdVnqi2Hbbbbnttts2Sj/nnHPy3pYQlcLZZ8Puu/u9B9ttV9i2pAQJIYQQoqwYPtzHHSs0Wg4TQgghRJdEM0ECnHRh0fVwzmFmpRajrHGldOUrRBHQ008I0eWor69n6dKlesgn4Jxj6dKl1NfXl1oUIQqGZoKEEF2Ofv36sXDhQj4ql1DWZUp9fT39+vUrtRhCFAwpQUKILkdtba08MgshtBwmhBBCiK6JlCAhhBBCdEmkBAkhhBCiS2LaHSGE6MyY2UfAO0Vqrg9QhIhHkqFCZIDykKMryjDAOde3vYukBAkhRJ4ws+eccyMkg2QoJzkkQzxaDhNCCCFEl0RKkBBCCCG6JFKChBAifzSWWgAkQ4pykAHKQw7JEINsgoQQQgjRJdFMkBBCCCG6JFKChBBCCNElkRIkhBAdwMz+z8xeN7M5ZnanmW0WpA80s1Vm9mJwXFNgOQ4zs3lm9paZfb+QbaW1u72ZPWxmr5rZK2Z2QZB+iZm9H+r/EQWWY4GZvRy09VyQtoWZ/d3M3gxeNy9g+0NCfX3RzD41s0nFGAczu97M/m1mc0NpkX03z6+Dz8kcM9ujgDKUxXcjUW7ZBAkhRO6Y2SHAQ8659WZ2GYBz7ntmNhD4q3Nu1yLIUA28AXwJWAg8C4xxzr1ahLa3AbZxzj1vZr2A2cCxwEnAcufc5YWWIZBjATDCObcklPZL4GPn3P8GiuHmzrnvFUGWauB9YG9gHAUeBzPbD1gO3Jj6vMX1PVDCvg4cEch3lXNu7wLJUPLvRntoJkgIITqAc+5B59z64PQpoF8JxNgLeMs59y/n3FrgVuCYYjTsnPvQOfd88H4Z8BqwXTHazoBjgBnB+xl45awYHATMd84VxVO5c+5R4OO05Li+H4NXVJxz7ilgs0CRzbsMZfLdSERKkBBC5I8zgftC54PM7AUze8TMRhew3e2A90LnCymBIhL8w98deDpIOj9YCrm+kEtRAQ540Mxmm9n4IG0r59yHwftFwFYFliHFycAtofNijkOKuL6X6rNSqu9GIlKChBCiHcxsppnNjTiOCV0zGVgPNAVJHwL9nXO7A98C/mBmmxRf+uJgZj2B24FJzrlPgWnAYGAYfiyuKLAI+zrn9gAOB/4nWJ7ZgPO2HwW3/zCzOuDLwJ+CpGKPw0YUq+9xlPN3o6YUjQohRCXhnDs4Kd/MzgCOAg4KHjg459YAa4L3s81sPvAZ4LkCiPg+sH3ovF+QVhTMrBavADU55+4AcM4tDuVfB/y1kDI4594PXv9tZnfilwgXm9k2zrkPgyWffxdShoDDgedT/S/2OISI63tRPytl8N1IRDNBQgjRAczsMOC7wJedcytD6X0DA1nMbAdgJ+BfBRLjWWAnMxsUzEScDNxdoLbaYGYG/B54zTl3ZSg9bGdyHDA3vWweZegRGGVjZj2AQ4L27gZODy47HbirUDKEGENoKayY45BGXN/vBk4LdomNBD4JLZvllTL5biTLqN1hQgiRO2b2FtANWBokPeWcO8/MvgJcCqwDWoCLnXP3FFCOI4BfAdXA9c65KYVqK63dfYHHgJfx/QT4AV4ZGIZfhlkAnFvAh+0OwJ3BaQ3wB+fcFDPrDdwG9AfeAU5yzqUbEOdTjh7Au8AOzrlPgrSbKPA4mNktwP5AH2AxcDHwFyL6HiitvwUOA1YC45xzHZ6BiZHhIsrgu5GElCAhhBBCdEm0HCaEEEKILomUICGEEEJ0SaQECSGEEKJLIiVICCGEEF0SKUFCCCGE6JJICRJCCCFEl0RKkBBCCCG6JFKChBBClCVmdqyZOTPbOeGa5cWUSXQupAQJIYQoV8bg40mNKbUgonMiJUgIIUTZEUSl3x84myyUIDMbaGZzQ+cXmtklQfrrZnaDmb1hZk1mdrCZPW5mb5rZXqEyfzGz2Wb2ipmNz2e/RHkhJUgIIUQ5cgww0zn3ErDczIbnoc4dgSuAnYPjFGBf4EJ8vLMUZzrnhgMjgG8EMchEJ0RKkBBCiHJkDD4AKMFrPpbE3nbOveycawFeAf7hfADNl4GBoeu+YWYvAU8B2+OjnItOiJQgIYQQZYX9//buGCWCIIqi6H0gKCaCA25jQjUQDAwNBXEHrsMFuBAXIBMbmhiqiRgoYmQmDMh8g+6gnUQnUqfuiao+XV2V9YNPU8kmsA1M+tIFcNzfgP6dD75+29YG4+lgPBvMZ3S3z5NkHzgAdqtqDNzMvUNLxBAkSfprjoDLqpoCVNUD8ALs/WDtK7CVZJRkFThccO8N4K2q3vu/0nYWXK9/ZOW3DyBJ0pwTYJzkcVAb9fWruWfXkzwN5ufAGXANPAN3C+49AU6T3AL3dC0xLal07VBJkqS22A6TJElNMgRJkqQmGYIkSVKTDEGSJKlJhiBJktQkQ5AkSWqSIUiSJDXpE5YYIYasKZfhAAAAAElFTkSuQmCC\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x7f9deda50b38>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"plt.figure(figsize=(10,4))\n",
"plt.subplot(1,2,1)\n",
"plt.imshow(im)\n",
"plt.title(\"Size: %d x %d\" % (block_size, block_size))\n",
"plt.axis('off');\n",
"\n",
"dy = np.ones(block_size) * round(np.mean(yuv[:,:,0])) - yuv[:,:,0]\n",
"du = np.ones(block_size) * round(np.mean(yuv[:,:,1])) - yuv[:,:,1]\n",
"dv = np.ones(block_size) * round(np.mean(yuv[:,:,2])) - yuv[:,:,2]\n",
"\n",
"plt.subplot(1,2,2)\n",
"plt.scatter(dy, du, color=\"b\")\n",
"plt.scatter(dy, dv, color=\"r\")\n",
"sdy = np.sort(dy.flatten())\n",
"plt.plot(sdy, sdy*best_a_cb, color=\"b\")\n",
"plt.plot(sdy, sdy*best_a_cr, color=\"r\")\n",
"plt.legend(['Pred Cb','Pred Cr', 'Cb', 'Cr'])\n",
"plt.title(\"Correlation between Luma and Chroma\")\n",
"plt.xlabel(\"$\\Delta$ Luma\")\n",
"plt.ylabel(\"$\\Delta$ Chroma\");"
]
},
{
"cell_type": "code",
"execution_count": 82,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x7f9dedabd358>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"plt.figure(figsize=(5,4))\n",
"plt.imshow(rgb)\n",
"plt.axis('off');"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.5.2"
}
},
"nbformat": 4,
"nbformat_minor": 1
}
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment