🚀 Image Segmentation & Thresholding - Assignment 2

This repository contains the solution for Take Home Assignment 2 of the Computer Vision and Image Processing (EC7212).

---

✨ Assignment Details

• Name: Perera G.A.L.S.
• Reg No: EG/2020/4111
• Semester: 07
• Course: EC7212 – Computer Vision and Image Processing
• Assignment: Take Home Assignment 2

---

🎯 Contents

✅ Question 1 – Otsu’s Thresholding

• A synthetic image with two objects (one circle, one rectangle) and three distinct pixel intensities (one for each object and one for the background) was generated using NumPy.
• Gaussian noise (mean = 0, standard deviation = 20) was added to the synthetic image.
• Otsu’s thresholding algorithm was applied using both OpenCV's built-in cv2.THRESH_OTSU method and scikit-image's filters.threshold_otsu.
• Outputs: Original image, noisy image, and the thresholded binary images from both OpenCV and scikit-image.

✅ Question 2 – Region Growing Segmentation

• A grayscale image was loaded using OpenCV.
• Seed points were manually selected within the region of interest.
• A custom region growing algorithm was implemented to segment regions based on the similarity of pixel intensities to the initial seed pixel values.
• The final output is a binary mask showcasing the segmented region.

---

🛠️ Requirements

• Python 3.x
• OpenCV (cv2)
• NumPy
• Matplotlib (for visualization)
• Scikit-image (skimage)

You can install the dependencies via:

pip install opencv-python numpy matplotlib scikit-image