atom-predict/egnn_v2/.ipynb_checkpoints/OTSUVis-checkpoint.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "2d4f51fc-c3e6-436f-90bb-849e54a62ee5",
   "metadata": {},
   "outputs": [],
   "source": [
    "import cv2\n",
    "import json\n",
    "import copy\n",
    "import numpy as np\n",
    "from PIL import Image\n",
    "import matplotlib.pyplot as plt\n",
    "from skimage.measure import label\n",
    "from skimage.measure import regionprops"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "246b2cde-704f-4930-b0c7-a1bb07d0b490",
   "metadata": {},
   "outputs": [],
   "source": [
    "json_path = '../msunet/logs/0/version_0/test.json'\n",
    "\n",
    "with open(json_path) as f:\n",
    "    data = json.load(f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "e3b627c6-67cc-4ea5-990f-74558b609c2c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(8, 2048, 2048)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pred = np.array(data['pred']); pred.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "59ee302c-4b42-45c0-b99b-cd4f37d5ce4b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['../../data/patch_unet//test/img/10.png',\n",
       " '../../data/patch_unet//test/img/4.png',\n",
       " '../../data/patch_unet//test/img/40.png',\n",
       " '../../data/patch_unet//test/img/20.png',\n",
       " '../../data/patch_unet//test/img/6.png',\n",
       " '../../data/patch_unet//test/img/8.png',\n",
       " '../../data/patch_unet//test/img/2.png',\n",
       " '../../data/patch_unet//test/img/30.png']"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data['img_path']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "1a52ef4e-443e-4a40-be32-201617b20fd8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 600x600 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(6, 6))\n",
    "patch = copy.deepcopy(pred[3][759:817, 540:600])\n",
    "patch[:19, 56:] = 0\n",
    "patch = cv2.dilate(patch, np.ones((3, 3)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "589524d8-6c4d-47dd-968a-fab14adbd9c7",
   "metadata": {},
   "outputs": [],
   "source": [
    "img = np.zeros((64, 64))\n",
    "img[3:61, 2:62] = patch\n",
    "patch = img"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "bf7c7fa2-789b-4c19-8ad0-3d2e15d8407d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(-0.5, 63.5, 63.5, -0.5)"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(patch, cmap='gray')\n",
    "plt.axis('off')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "657391f6-0846-4fda-b1d2-4d4a9e9fe472",
   "metadata": {},
   "outputs": [],
   "source": [
    "binary = np.array(patch * 255., np.uint8)\n",
    "_, binary = cv2.threshold(binary, 0, 255, cv2.THRESH_BINARY + cv2.THRESH_OTSU)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "c11519a3-2816-4b3e-abf8-f0215a7f6d0a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(-0.5, 63.5, 63.5, -0.5)"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(binary, cmap='gray')\n",
    "plt.axis('off')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "151c330e-6274-4021-97d7-2273cb1ddce9",
   "metadata": {},
   "outputs": [],
   "source": [
    "centers = []\n",
    "mask = np.zeros(binary.shape)\n",
    "\n",
    "labeled_img = label(binary)\n",
    "rps = regionprops(labeled_img)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "9a5c998a-9747-4c41-9bf3-88afd6afa307",
   "metadata": {},
   "outputs": [],
   "source": [
    "def hough_center_detection(i, rp, labeled_img, img_size=2048):\n",
    "    hs, ws, he, we = rp.bbox\n",
    "    hs = np.clip(hs - 8, 0, img_size-1)\n",
    "    ws = np.clip(ws - 8, 0, img_size-1)\n",
    "    he = np.clip(he + 8, 0, img_size-1)\n",
    "    we = np.clip(we + 8, 0, img_size-1)\n",
    "\n",
    "    m = np.array(labeled_img == rp.label, np.uint8)[hs:he, ws:we]\n",
    "    m = cv2.dilate(m, np.ones((3, 3)))  \n",
    "    \n",
    "    cricles = cv2.HoughCircles(\n",
    "        m,\n",
    "        method = cv2.HOUGH_GRADIENT,\n",
    "        dp = 1,\n",
    "        minDist = 13,\n",
    "        minRadius = 5,\n",
    "        maxRadius = 12,\n",
    "        param1 = 5,\n",
    "        param2 = 6,\n",
    "    )\n",
    "    \n",
    "    if cricles is None:\n",
    "        return np.array([])\n",
    "    \n",
    "    if (rp.area > 400) & (cricles.shape[1] != 2):\n",
    "        print(i, cricles.shape[1])\n",
    "        \n",
    "    print(cricles[0], cricles[0][:, :2][:, ::-1] + [hs, ws])\n",
    "    \n",
    "    centers = np.round(cricles[0][:, :2][:, ::-1] + [hs, ws])\n",
    "    centers = np.array(centers, np.int32)\n",
    "    \n",
    "    return centers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "f18c9429-0f1e-4f59-8207-4ea9b5d43db9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[16.5 12.5 10.3]\n",
      " [31.5 18.5 11.4]] [[12.5 31.5]\n",
      " [18.5 46.5]]\n",
      "[[ 9.5 17.5 10.2]] [[17.5  9.5]]\n",
      "[[16.5 15.5  9.6]] [[36.5 51.5]]\n",
      "[[13.5 18.5 11.4]] [[41.5 13.5]]\n",
      "[[15.5 16.5 10.3]] [[51.5 31.5]]\n"
     ]
    }
   ],
   "source": [
    "for i, rp in enumerate(rps):\n",
    "    if rp.area < 48:\n",
    "        continue\n",
    "\n",
    "    rp_centers = hough_center_detection(i, rp, labeled_img)\n",
    "\n",
    "    if len(rp_centers) == 0:\n",
    "        h, w = np.array(np.round(rp.centroid), np.int32)\n",
    "        mask[h, w] = 1\n",
    "    else:\n",
    "        for h, w in rp_centers:\n",
    "            mask[h, w] = 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "e0637ceb-1c56-4352-a56c-eaf1dbf9938c",
   "metadata": {},
   "outputs": [],
   "source": [
    "h, w = np.array([12, 18, 18, 36, 42, 52]), np.array([32, 10, 46, 52, 14, 32])\n",
    "r = np.round(np.array([10.3, 11.4, 10.2, 9.6, 11.4, 10.3]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "acc6a59f-db70-4532-80c6-4993fcc262a7",
   "metadata": {},
   "outputs": [],
   "source": [
    "from PIL import Image"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "d95be2e4-122f-436a-9907-33495556afc1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(-0.5, 63.5, 63.5, -0.5)"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "bbrgb = np.array(Image.fromarray(binary).convert('RGB'))\n",
    "\n",
    "for i in range(len(h)):\n",
    "    cv2.circle(bbrgb, (w[i], h[i]), int(r[i]), [255, 0, 0])\n",
    "    plt.scatter(w[i], h[i])\n",
    "    \n",
    "plt.imshow(bbrgb, cmap='gray')\n",
    "plt.axis('off')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "58ebe6a1-62fd-4188-8da1-c7758b6c6978",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4104cebe-15cf-4b67-97da-91cd9f5c9ee1",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "cmae",
   "language": "python",
   "name": "cmae"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.16"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}