lite: explicitly install opencv-python in Pyodide setup cell

Author: petercorke

lite/files/intro.ipynb

@@ -0,0 +1,236 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "id": "0f90fd53",
+   "metadata": {},
+   "source": [
+    "# Machine Vision Toolbox for Python — Interactive Demo\n",
+    "\n",
+    "This notebook runs entirely in your browser using [JupyterLite](https://jupyterlite.readthedocs.io) and [Pyodide](https://pyodide.org).\n",
+    "No installation required.  The cell below installs the toolbox the first time it is run (takes ~30 s)."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "c663dd60",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import sys\n",
+    "\n",
+    "if sys.platform == \"emscripten\":\n",
+    "    import micropip\n",
+    "\n",
+    "    # Install all dependencies explicitly.\n",
+    "    # opencv-python, numpy, scipy and matplotlib are in the Pyodide package\n",
+    "    # index but must be loaded explicitly; they are NOT auto-loaded.\n",
+    "    await micropip.install([\n",
+    "        \"opencv-python\",\n",
+    "        \"spatialmath-python\",\n",
+    "        \"pgraph-python\",\n",
+    "        \"ansitable\",\n",
+    "        \"mvtb-data\",\n",
+    "    ])\n",
+    "\n",
+    "    # Install the toolbox without re-resolving compiled deps already loaded above.\n",
+    "    await micropip.install(\"machinevision-toolbox-python\", deps=False)\n",
+    "\n",
+    "print(\"Ready.\")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "4bd104d5",
+   "metadata": {},
+   "source": [
+    "## Images and pixels\n",
+    "\n",
+    "The core class is `Image`.  Let's read one of the bundled images and inspect it."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "4fc8b259",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import matplotlib\n",
+    "matplotlib.use(\"module://matplotlib_pyodide.html5_canvas_backend\")\n",
+    "import matplotlib.pyplot as plt\n",
+    "\n",
+    "from machinevisiontoolbox import Image\n",
+    "\n",
+    "mona = Image.Read(\"monalisa.png\")\n",
+    "print(f\"size: {mona.width} x {mona.height}, dtype: {mona.dtype}, colour: {mona.iscolor}\")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "fe77ee73",
+   "metadata": {},
+   "source": [
+    "Display the image inline with `disp()`, which uses matplotlib."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "c342305d",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "mona.disp()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "43219088",
+   "metadata": {},
+   "source": [
+    "## Smoothing\n",
+    "\n",
+    "Apply a Gaussian blur and display original and result side-by-side."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "964566d0",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "smooth = mona.smooth(sigma=3)\n",
+    "\n",
+    "fig, axes = plt.subplots(1, 2, figsize=(10, 5))\n",
+    "mona.disp(ax=axes[0], title=\"Original\")\n",
+    "smooth.disp(ax=axes[1], title=\"Smoothed (sigma=3)\")\n",
+    "plt.tight_layout()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "88224949",
+   "metadata": {},
+   "source": [
+    "## Greyscale and histograms"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "6ebccd0a",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "grey = mona.mono()\n",
+    "print(f\"grey: {grey.width} x {grey.height}, planes: {grey.nplanes}\")\n",
+    "\n",
+    "hist, x = grey.hist()\n",
+    "plt.figure()\n",
+    "plt.bar(x, hist, width=1, color=\"steelblue\")\n",
+    "plt.xlabel(\"Pixel value\")\n",
+    "plt.ylabel(\"Count\")\n",
+    "plt.title(\"Greyscale histogram\")\n",
+    "plt.tight_layout()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "60a85be2",
+   "metadata": {},
+   "source": [
+    "## Colour planes\n",
+    "\n",
+    "Access individual colour planes by name — the toolbox tracks the colour order so\n",
+    "you never need to worry about BGR vs RGB."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "3af193e5",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "flowers = Image.Read(\"flowers1.png\")\n",
+    "print(f\"colour order: {flowers.colororder_str}\")\n",
+    "\n",
+    "fig, axes = plt.subplots(1, 3, figsize=(12, 4))\n",
+    "for ax, plane in zip(axes, [\"R\", \"G\", \"B\"]):\n",
+    "    flowers.plane(plane).disp(ax=ax, title=plane, colormap=\"gray\")\n",
+    "plt.tight_layout()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "01cd03d5",
+   "metadata": {},
+   "source": [
+    "## Binary blobs\n",
+    "\n",
+    "Load a binary image and find the blobs."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "fb808a02",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "sharks = Image.Read(\"shark2.png\")\n",
+    "blobs = sharks.blobs()\n",
+    "print(blobs)\n",
+    "\n",
+    "fig, ax = plt.subplots()\n",
+    "sharks.disp(ax=ax)\n",
+    "blobs.plot_box(ax, color=\"g\")\n",
+    "blobs.plot_centroid(ax, \"o\", color=\"y\")\n",
+    "plt.tight_layout()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "829820e5",
+   "metadata": {},
+   "source": [
+    "## Camera model\n",
+    "\n",
+    "Create a central perspective camera and project a 3-D point into the image plane."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "de23946b",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from machinevisiontoolbox import CentralCamera\n",
+    "from spatialmath import SE3\n",
+    "\n",
+    "cam = CentralCamera(f=0.015, rho=10e-6, imagesize=[1280, 1024],\n",
+    "                    pp=[640, 512], name=\"mycamera\")\n",
+    "print(cam)\n",
+    "\n",
+    "P = [0.3, 0.4, 3.0]\n",
+    "p = cam.project_point(P)\n",
+    "print(f\"Projected pixel: {p}\")\n",
+    "\n",
+    "# Shift camera 100 mm to the right\n",
+    "p2 = cam.project_point(P, pose=SE3(0.1, 0, 0))\n",
+    "print(f\"Projected pixel (camera shifted): {p2}\")"
+   ]
+  }
+ ],
+ "metadata": {
+  "language_info": {
+   "name": "python"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}