pixel_math.ipynb

{
 "cells": [
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": "# Pixel Math\n\nA visual reminder of pixel math."
  },
  {
   "metadata": {
    "trusted": true,
    "collapsed": false
   },
   "cell_type": "code",
   "source": "import numpy as np\n\n# Make a 2x2 array of R,G,B pixels (or BGR for CV2)\npixel00 = [255,255,255]\npixel01 = [200,200,200]\npixel10 = [0,0,0]\npixel11 = [100,100,100]\nd = np.array([[pixel00, pixel01],[pixel10,pixel11]])",
   "execution_count": 21,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true,
    "collapsed": false
   },
   "cell_type": "code",
   "source": "# Confirm the shape\nd.shape",
   "execution_count": 22,
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": "(2, 2, 3)"
     },
     "metadata": {},
     "execution_count": 22
    }
   ]
  },
  {
   "metadata": {
    "trusted": true,
    "collapsed": false
   },
   "cell_type": "code",
   "source": "# view the data\nd",
   "execution_count": 23,
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": "array([[[255, 255, 255],\n        [200, 200, 200]],\n\n       [[  0,   0,   0],\n        [100, 100, 100]]])"
     },
     "metadata": {},
     "execution_count": 23
    }
   ]
  },
  {
   "metadata": {
    "trusted": true,
    "collapsed": true
   },
   "cell_type": "code",
   "source": "# Now load a 32x32 RGB image from the cifar10 dataset\nimport cv2\nim = cv2.imread(\"airplane4.png\")",
   "execution_count": 24,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true,
    "collapsed": false
   },
   "cell_type": "code",
   "source": "im.shape",
   "execution_count": 25,
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": "(32, 32, 3)"
     },
     "metadata": {},
     "execution_count": 25
    }
   ]
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": "![title](https://user-images.githubusercontent.com/28526/47608637-a42dc000-d9e5-11e8-9f3b-4d73051d0db0.png)"
  },
  {
   "metadata": {
    "trusted": true,
    "collapsed": false
   },
   "cell_type": "code",
   "source": "# Looks similar to 2x2\nim[0]",
   "execution_count": 27,
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": "array([[177, 124,  57],\n       [170, 122,  57],\n       [174, 123,  57],\n       [179, 123,  59],\n       [178, 125,  62],\n       [178, 128,  61],\n       [179, 129,  62],\n       [178, 129,  61],\n       [180, 132,  59],\n       [183, 132,  61],\n       [183, 132,  61],\n       [184, 133,  62],\n       [184, 133,  62],\n       [185, 134,  63],\n       [185, 134,  63],\n       [185, 134,  63],\n       [186, 136,  64],\n       [187, 136,  65],\n       [188, 135,  65],\n       [188, 135,  64],\n       [188, 136,  64],\n       [189, 138,  63],\n       [190, 139,  61],\n       [190, 137,  64],\n       [190, 137,  64],\n       [189, 136,  64],\n       [184, 134,  67],\n       [182, 134,  67],\n       [183, 134,  67],\n       [184, 131,  69],\n       [189, 131,  65],\n       [186, 132,  63]], dtype=uint8)"
     },
     "metadata": {},
     "execution_count": 27
    }
   ]
  },
  {
   "metadata": {
    "trusted": true,
    "collapsed": false
   },
   "cell_type": "code",
   "source": "# to normalize for feeding into a ML net - e.g. neural network\nd = d / 255.0",
   "execution_count": 9,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true,
    "collapsed": false
   },
   "cell_type": "code",
   "source": "d",
   "execution_count": 11,
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": "array([[[1.        , 1.        , 1.        ],\n        [0.78431373, 0.78431373, 0.78431373]],\n\n       [[0.        , 0.        , 0.        ],\n        [0.39215686, 0.39215686, 0.39215686]]])"
     },
     "metadata": {},
     "execution_count": 11
    }
   ]
  },
  {
   "metadata": {
    "trusted": true,
    "collapsed": false
   },
   "cell_type": "code",
   "source": "# reshape to feed as single feature vector\nd.reshape(d.shape[0] * d.shape[1] * d.shape[2])",
   "execution_count": 20,
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": "array([1.        , 1.        , 1.        , 0.78431373, 0.78431373,\n       0.78431373, 0.        , 0.        , 0.        , 0.39215686,\n       0.39215686, 0.39215686])"
     },
     "metadata": {},
     "execution_count": 20
    }
   ]
  },
  {
   "metadata": {
    "trusted": true,
    "collapsed": true
   },
   "cell_type": "code",
   "source": "",
   "execution_count": null,
   "outputs": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "name": "conda-env-ocr-py",
   "display_name": "Python [conda env:ocr]",
   "language": "python"
  },
  "language_info": {
   "nbconvert_exporter": "python",
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "name": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.4",
   "mimetype": "text/x-python",
   "file_extension": ".py"
  },
  "gist_id": "618c65e98b70261c616ada422f09373d"
 },
 "nbformat": 4,
 "nbformat_minor": 0
}

Link to image: