Deforestation Segmentation Using Raw Image Processing in MATLAB

This project analyzes deforestation over time by comparing satellite images from 2000 and 2017. It leverages MATLAB for image processing, segmentation, and area calculation using color thresholding techniques.

---

Features

• Image Comparison: Display and compare images from 2000 and 2017.
• Grayscale Conversion: Convert original images to grayscale.
• Intensity Adjustment: Adjust grayscale intensity for improved segmentation.
• Color Segmentation: Use color thresholding to segment deforested regions.
• Area Calculation: Compute the area of deforested regions in pixels and convert them to square kilometers.

---

Project Structure

• MLX File: deforestation_segmentation.mlx
  Contains the main script for reading images, processing them, and performing segmentation.

• M File: createMask.m
  Auto-generated mask function from the MATLAB Color Thresholder app, used to segment regions based on color.

• Images:

  • 2000.jpg: Satellite image from the year 2000.
  • 2017.jpg: Satellite image from the year 2017.
  • (Include any additional images if applicable)

---

Dependencies

• MATLAB: R2022a or later recommended.
• Image Processing Toolbox: Required for image adjustments and segmentation.

---

How to Run

1. Open the project folder in MATLAB.
2. Run deforestation_segmentation.mlx to visualize the comparison and segmentation.
3. Use the createMask function for additional custom segmentation, if needed.

---

Area Calculation Formula

To convert pixels to square kilometers:

• Given scale: 20 km = 58 px

• 1 px = ( \frac{20}{58} ) km

• Area of 1 px = ( \left( \frac{20}{58} \right)^2 ) km²

• The total area is computed using:

    px2km = (20 / 58)^2;
    area2000 = round(areaPx2000 * px2km);
    area2017 = round(areaPx2017 * px2km);
  

Results

The segmented output displays the deforested regions for both years:

• Year 2000: Area = area2000 km²

• Year 2017: Area = area2017 km²