notebook.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Building a Motorized Maze Rig"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Importing all required libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import cv2 as cv\n",
    "from matplotlib import pyplot as plt\n",
    "from ipywidgets import interactive\n",
    "import numpy as np"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load image and calibrate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "img = cv.imread('images/maze01.jpg')\n",
    "img = cv.cvtColor(img, cv.COLOR_BGR2RGB)\n",
    "plt.imshow(img)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![Calibration image](./images/my_photo01.jpg)\n",
    "![Calibration image](./images/my_photo05.jpg)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "calibration = np.load('rig/calib.npz')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "mtx = calibration['mtx']\n",
    "dist = calibration['dist']\n",
    "rvecs = calibration['rvecs']\n",
    "tvecs = calibration['tvecs']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "h, w = img.shape[:2]"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Create camera matrix for calibration."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "newCameraMtx, roi = cv.getOptimalNewCameraMatrix(mtx, dist, (w, h), 1, (w, h))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "undistortedImg = cv.undistort(img, mtx, dist, None, newCameraMtx)\n",
    "plt.imshow(undistortedImg)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Find corner HSV range"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "hsv_img = cv.cvtColor(undistortedImg, cv.COLOR_RGB2HSV)\n",
    "def mask(h, sl, vl, sh, vh):\n",
    "  m = cv.inRange(\n",
    "    hsv_img,\n",
    "    (h - 10, sl, vl),\n",
    "    (h + 10, sh, vh),\n",
    "  )\n",
    "  plt.imshow(m)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "interactive(\n",
    "  mask,\n",
    "  h=(0, 179),\n",
    "  sl=(0, 255),\n",
    "  vl=(0, 255),\n",
    "  sh=(0, 255),\n",
    "  vh=(0, 255),\n",
    ")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Create mask for corners"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# TODO: Change to found values\n",
    "CORNER_LOWER = (0, 0, 0)\n",
    "CORNER_UPPER = (179, 255, 255)\n",
    "\n",
    "# TODO: Create mask for corners"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Find corners in ROI"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Find corners in regions of interest.\n",
    "\n",
    "![Corners](./images/corners_img.png)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Fix perspective of the image"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "By fixing image perspective we both crop it and change the perspective so it is basically a matrix which can be easily parsed."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Map the maze"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Once we have an image with fixed perspective, we can create a map of the maze. This could be a simple matrix whose values are classes of the fields of the maze. Last thing to do before we solve the maze is to classify each field."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Solving the maze"
   ]
  }
 ],
 "metadata": {
  "language_info": {
   "name": "python"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}