{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Segment Anything Model for Geospatial Data \n",
    "\n",
    "[![image](https://studiolab.sagemaker.aws/studiolab.svg)](https://studiolab.sagemaker.aws/import/github/opengeos/segment-geospatial/blob/main/docs/examples/satellite-predictor.ipynb)\n",
    "[![image](https://img.shields.io/badge/Open-Planetary%20Computer-black?style=flat&logo=microsoft)](https://pccompute.westeurope.cloudapp.azure.com/compute/hub/user-redirect/git-pull?repo=https://github.com/opengeos/segment-geospatial&urlpath=lab/tree/segment-geospatial/docs/examples/satellite-predictor.ipynb&branch=main)\n",
    "[![image](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/opengeos/segment-geospatial/blob/main/docs/examples/satellite-predictor.ipynb)\n",
    "\n",
    "This notebook shows how to use segment satellite imagery using the Segment Anything Model (SAM) with a few lines of code. \n",
    "\n",
    "Make sure you use GPU runtime for this notebook. For Google Colab, go to `Runtime` -> `Change runtime type` and select `GPU` as the hardware accelerator. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Install dependencies\n",
    "\n",
    "Uncomment and run the following cell to install the required dependencies.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "outputId": "6924c5c8-39a0-4ac6-f114-9f1f8d102e88"
   },
   "outputs": [],
   "source": [
    "# %pip install segment-geospatial"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Import libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "import os\n",
    "import leafmap\n",
    "from samgeo import SamGeoPredictor, tms_to_geotiff, get_basemaps\n",
    "from segment_anything import sam_model_registry"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Create an interactive map"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "zoom = 16\n",
    "m = leafmap.Map(center=[45, -123], zoom=zoom)\n",
    "m.add_basemap(\"SATELLITE\")\n",
    "m"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Pan and zoom the map to select the area of interest. Use the draw tools to draw a polygon or rectangle on the map"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "if m.user_roi_bounds() is not None:\n",
    "    bbox = m.user_roi_bounds()\n",
    "else:\n",
    "    bbox = [-123.0127, 44.9957, -122.9874, 45.0045]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Download map tiles\n",
    "\n",
    "Download maps tiles and mosaic them into a single GeoTIFF file"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "image = \"satellite.tif\"\n",
    "# image = '/path/to/your/own/image.tif'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Besides the `satellite` basemap, you can use any of the following basemaps returned by the `get_basemaps()` function:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# get_basemaps().keys()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Specify the basemap as the source."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "tms_to_geotiff(\n",
    "    output=image, bbox=bbox, zoom=zoom + 1, source=\"Satellite\", overwrite=True\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "m.add_raster(image, layer_name=\"Image\")\n",
    "m"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "tags": []
   },
   "source": [
    "Use the draw tools to draw a rectangle from which to subset segmentations on the map"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "if m.user_roi_bounds() is not None:\n",
    "    clip_box = m.user_roi_bounds()\n",
    "else:\n",
    "    clip_box = [-123.0064, 44.9988, -123.0005, 45.0025]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "clip_box"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Initialize SamGeoPredictor class"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "out_dir = os.path.join(os.path.expanduser(\"~\"), \"Downloads\")\n",
    "checkpoint = os.path.join(out_dir, \"sam_vit_h_4b8939.pth\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "import cv2\n",
    "\n",
    "img_arr = cv2.imread(image)\n",
    "\n",
    "model_type = \"vit_h\"\n",
    "\n",
    "sam = sam_model_registry[model_type](checkpoint=checkpoint)\n",
    "\n",
    "predictor = SamGeoPredictor(sam)\n",
    "\n",
    "predictor.set_image(img_arr)\n",
    "\n",
    "masks, _, _ = predictor.predict(src_fp=image, geo_box=clip_box)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "masks_img = \"preds.tif\"\n",
    "predictor.masks_to_geotiff(image, masks_img, masks.astype(\"uint8\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "vector = \"feats.geojson\"\n",
    "gdf = predictor.geotiff_to_geojson(masks_img, vector, bidx=1)\n",
    "gdf.plot()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Visualize the results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "style = {\n",
    "    \"color\": \"#3388ff\",\n",
    "    \"weight\": 2,\n",
    "    \"fillColor\": \"#7c4185\",\n",
    "    \"fillOpacity\": 0.5,\n",
    "}\n",
    "m.add_vector(vector, layer_name=\"Vector\", style=style)\n",
    "m"
   ]
  }
 ],
 "metadata": {
  "accelerator": "GPU",
  "colab": {
   "provenance": []
  },
  "gpuClass": "standard",
  "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": "3.9.16"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}