Loading a subset of images to copy them in the data directory of SegCaps

LoadingDataset-Second Generation-save for SegCaps.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "#import h5py\n",
    "import numpy as np\n",
    "from matplotlib import pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.13.1\n"
     ]
    }
   ],
   "source": [
    "import skimage as sk\n",
    "print(sk.__version__)\n",
    "from skimage import io\n",
    "#from skimage import filters\n",
    "#from skimage import feature\n",
    "\n",
    "#from scipy import ndimage as nd\n",
    "#from scipy import misc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cy3.tif\n",
      "DAPI.tif\n",
      "FullRes\n",
      "LowRes\n",
      "LowRes_13434_overlapping_pairs.h5\n",
      "lowres_82146_overlapping_pairs_grey_DAPI-GroundTruth.h5\n",
      "overlapping_chromosomes_examples.h5\n",
      "overlapping_subset_pairs.h5\n",
      "\n",
      "Second generation of low resolution dataset:\n",
      "README.md\n",
      "train\n",
      "validation\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from subprocess import check_output\n",
    "print(check_output([\"ls\", \"../dataset\"]).decode(\"utf8\"))\n",
    "print('Second generation of low resolution dataset:')\n",
    "print(check_output([\"ls\", \"../dataset/LowRes\"]).decode(\"utf8\"))"
   ]
  },
  {
   "attachments": {
    "pair-grey_mask.png": {
     "image/png": "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"
    }
   },
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Load images from second generation dataset\n",
    "\n",
    "The dataset of low resolution images is available on github:\n",
    "https://github.com/jeanpat/DeepFISH/tree/png-dataset\n",
    "   * Here the images are stored as pairs of png images in two directories of the data/ in the SegCaps project\n",
    "       * grey\n",
    "       * groundtruth (multilevel masks)\n",
    "       \n",
    "       ![pair-grey_mask.png](attachment:pair-grey_mask.png)\n",
    "       \n",
    "Then copy a subset of the images in the *_data/_* directory of SegCaps:\n",
    "https://github.com/Cheng-Lin-Li/SegCaps"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "100085  images found\n",
      "../dataset/LowRes/train/grey/grey0000280.png - ../dataset/LowRes/train/groundtruth/gtruth0000280.png\n"
     ]
    }
   ],
   "source": [
    "top = os.path.join('..','dataset','LowRes','train','grey')\n",
    "path, dirs, files = next(os.walk(top))\n",
    "file_count = len(files)\n",
    "print(file_count,' images found')\n",
    "N_im = 100 # take a subset of images\n",
    "idx_images = np.random.randint(1,file_count, N_im)#numbers of pairs of images to load\n",
    "idx = []\n",
    "for n in idx_images:\n",
    "    idx.append('%07d' % n)\n",
    "greys_id = []\n",
    "masks_id = []\n",
    "for i in idx:\n",
    "    greys_id.append(os.path.join(top,'grey'+i+'.png'))\n",
    "    masks_id.append(os.path.join('..','dataset','LowRes','train','groundtruth','gtruth'+i+'.png'))\n",
    "#grey0066893.png\n",
    "#gtruth0075056.png \n",
    "greys_id.sort()\n",
    "masks_id.sort()\n",
    "print(greys_id[0], '-',masks_id[0])\n",
    "grey_col = sk.io.ImageCollection(greys_id)\n",
    "mask_col = sk.io.ImageCollection(masks_id)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7f4758170898>"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x864 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "N = 99\n",
    "plt.figure(figsize=(8,12))\n",
    "plt.subplot(121)\n",
    "plt.title('One channel greyscaled')\n",
    "plt.imshow(grey_col[N], cmap=plt.cm.gist_gray)\n",
    "plt.subplot(122)\n",
    "plt.title('Mask: background=0')\n",
    "plt.imshow(mask_col[N], cmap=plt.cm.jet)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x864 with 100 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12,12))\n",
    "for k, im in enumerate(grey_col):\n",
    "    plt.subplot(10,10,k+1, xticks=[], yticks=[])\n",
    "    plt.imshow(im, cmap=plt.cm.gray)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x864 with 100 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12,12))\n",
    "for k, im in enumerate(mask_col):\n",
    "    plt.subplot(10,10,k+1, xticks=[], yticks=[])\n",
    "    plt.imshow(im, jet)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "# Save pairs of grey+masks in SegCaps/data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "grey_path_to_segcaps = os.path.join('..','..','SegCaps','data','imgs')\n",
    "mask_path_to_segcaps = os.path.join('..','..','SegCaps','data','masks')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'/home/jeanpat/Developpement/DeepFISH/notebooks'"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "os.path.abspath('.')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000000.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000001.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000002.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000003.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000004.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000005.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000006.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000007.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000008.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000009.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000010.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000011.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000012.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000013.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000014.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000015.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000016.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000017.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000018.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000019.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000020.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000021.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000022.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000023.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000024.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000025.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000026.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000027.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000028.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000029.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000030.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000031.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000032.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000033.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000034.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000035.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000036.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000037.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000038.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000039.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000040.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000041.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000042.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000043.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000044.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000045.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000046.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000047.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000048.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000049.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000050.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000051.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000052.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000053.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000054.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000055.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000056.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000057.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000058.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000059.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000060.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000061.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000062.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000063.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000064.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000065.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000066.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000067.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000068.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000069.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000070.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000071.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000072.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000073.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000074.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000075.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000076.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000077.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000078.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000079.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000080.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000081.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000082.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000083.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000084.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000085.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000086.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000087.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000088.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000089.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000090.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000091.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000092.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000093.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000094.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000095.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000096.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000097.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000098.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n",
      "/home/jeanpat/anaconda3/envs/DeepFish/lib/python3.6/site-packages/skimage/io/_io.py:132: UserWarning: ../../SegCaps/data/masks/train0000099.png is a low contrast image\n",
      "  warn('%s is a low contrast image' % fname)\n"
     ]
    }
   ],
   "source": [
    "for idx,grey_mask in enumerate(zip(grey_col, mask_col)):\n",
    "    grey = grey_mask[0]\n",
    "    mask = grey_mask[1]\n",
    "    im_idx = '%07d' % idx\n",
    "    io.imsave(os.path.join(grey_path_to_segcaps,'train'+im_idx+'.png'), grey)\n",
    "    io.imsave(os.path.join(mask_path_to_segcaps,'train'+im_idx+'.png'), mask)"
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "python3 ./main.py --test --Kfold 2 --net segcapsr3 --data_root_dir=data --loglevel 2 --which_gpus=-2 --gpus=0 --dataset mscoco17 --weights_path data/saved_models/segcapsr3/split-0_batch-1_shuff-1_aug-1_loss-dice_slic-1_sub--1_strid-1_lr-0.01_recon-2.0_model_20180702-055808.hdf5"
   ]
  }
 ],
 "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": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}