{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Binarization\n",
    "\n",
    "## 0. Import package"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "\n",
    "import cv2\n",
    "import collections\n",
    "\n",
    "import openalea.phenomenal.image as phm_img\n",
    "import openalea.phenomenal.data as phm_data\n",
    "import openalea.phenomenal.display as phm_display"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1. Prerequisites\n",
    "\n",
    "### 1. Load images\n",
    "\n",
    "Here, we load all the image of the dataset to simply the toturial, commodly you load just the images you want process one after the other."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "raw_images = phm_data.raw_images(\"data/plant_6\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1.3. Display images \n",
    "\n",
    "You can view each image according to this date and angle of view like this :\n",
    "Note : Angle top view is represented by negative number (-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "phm_display.show_images(\n",
    "    [raw_images[\"side\"][120], raw_images[\"side\"][30], raw_images[\"top\"][0]]\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2. Binarization\n",
    "\n",
    "### 2.1. Define a binarization routines"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def routine_side_binarization(image, mean_img):\n",
    "    maks = phm_data.tutorial_data_binarization_mask(\"data\")\n",
    "\n",
    "    threshold = 0.3\n",
    "    dark_background = False\n",
    "\n",
    "    hsv_min = (30, 11, 0)\n",
    "    hsv_max = (129, 254, 141)\n",
    "\n",
    "    # Convert image on HSV representation\n",
    "    hsv_image = cv2.cvtColor(image, cv2.COLOR_BGR2HSV)\n",
    "    # Threshold the image with HSV min and max value\n",
    "    binary_hsv_image = phm_img.threshold_hsv(hsv_image, hsv_min, hsv_max, maks[0])\n",
    "\n",
    "    # Threshold the image with difference between image and mean_image\n",
    "    binary_mean_shift_image = phm_img.threshold_meanshift(\n",
    "        image, mean_img, threshold, dark_background, maks[1]\n",
    "    )\n",
    "\n",
    "    # Add the two image\n",
    "    result = cv2.add(binary_hsv_image, binary_mean_shift_image)\n",
    "\n",
    "    # Erode and dilate the image to remove possible noise\n",
    "    result = cv2.medianBlur(result, 3)\n",
    "\n",
    "    return result\n",
    "\n",
    "\n",
    "def routine_top_binarization(image):\n",
    "    hsv_min = (42, 75, 28)\n",
    "    hsv_max = (80, 250, 134)\n",
    "    median_blur_size = 9\n",
    "    iterations = 5\n",
    "\n",
    "    # Convert image on HSV representation\n",
    "    hsv_image = cv2.cvtColor(image, cv2.COLOR_BGR2HSV)\n",
    "    # Apply a median blur on the image\n",
    "    hsv_image = cv2.medianBlur(hsv_image, ksize=median_blur_size)\n",
    "\n",
    "    # Threshold the image with HSV min and max value\n",
    "    bin_img = phm_img.threshold_hsv(hsv_image, hsv_min, hsv_max)\n",
    "    # dilate and erode the image to remove possible noise\n",
    "    bin_img = phm_img.dilate_erode(bin_img, kernel_shape=(3, 3), iterations=iterations)\n",
    "\n",
    "    return bin_img"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2.2. Binarize images"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Compute the mean image of the side view image\n",
    "\n",
    "mean_img = phm_img.mean_image(list(raw_images[\"side\"].values()))\n",
    "\n",
    "routine_binarization = {\n",
    "    \"side\": lambda im: routine_side_binarization(im, mean_img),\n",
    "    \"top\": lambda im: routine_top_binarization(im),\n",
    "}\n",
    "\n",
    "bin_images = collections.defaultdict(dict)\n",
    "for id_camera in raw_images:\n",
    "    for angle in raw_images[id_camera]:\n",
    "        bin_images[id_camera][angle] = routine_binarization[id_camera](\n",
    "            raw_images[id_camera][angle]\n",
    "        )"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2.3. Display images binarize "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "id_camera, angle = \"side\", 120\n",
    "phm_display.show_images(\n",
    "    [raw_images[id_camera][angle], mean_img, bin_images[id_camera][angle]]\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "id_camera, angle = \"top\", 0\n",
    "phm_display.show_images([raw_images[id_camera][angle], bin_images[id_camera][angle]])"
   ]
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.12.7"
  },
  "widgets": {
   "application/vnd.jupyter.widget-state+json": {
    "state": {},
    "version_major": 2,
    "version_minor": 0
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}