computer_vision_uni_coursework

This repository contains three completed courseworks from the Computer Vision course at University of Manchester. Each coursework demonstrates different techniques and methodologies in computer vision, providing a comprehensive overview of the concepts learned throughout the course.

Courseworks Overview

Coursework 1: Image Smoothing and Edge Detection

• Objective: Apply smoothing filters and edge detection techniques.
• Key Techniques:
  • Average (Mean) Smoothing Filter
  • Weighted-Average (Gaussian) Smoothing Filter
  • Sobel Operator for Edge Detection
  • Thresholding for Object Segmentation

Coursework 2: Object Detection

• Objective: Detect objects in images using feature detection and matching.
• Key Techniques:
  • Harris Corner Detection
    • Feature Detection
    • Implementation from Scratch
  • ORB (Oriented FAST and Rotated BRIEF) Feature Detection and Description
    • Built-in ORB Framework
    • Keypoint Detection
  • Feature Matching using Sum-of-Squared Differences (SSD)
    • Implementation using SciPy
    • Distance Calculation
  • Ratio Test for Vague Matches
    • Euclidean Distance
    • Thresholding for Valid Matches

Coursework 3: Stereo Imagery and Selective Focus

• Objective: Handle stereo imagery to create 3D models and apply selective focus.
• Key Techniques:
  • Focal Length Calculation
    • Formula Application
    • Parameter Estimation
  • Disparity Map Creation using Canny Edge Detection
    • Edge Detection with Canny
    • Disparity Calculation
  • 3D Scene Reconstruction from Disparity Data
    • Coordinate Calculation
    • 3D Visualization
  • Selective Focus based on Depth Information
    • Depth Image Creation
    • Foreground and Background Segmentation

Prerequisites

• Python 3.x
• NumPy
• OpenCV
• Matplotlib (for Coursework 3)