Module 11–12: Capstone Project

1. Overview

This two-week module is dedicated to your capstone project — an opportunity to integrate and apply everything you’ve learned in this course. You will define a computer vision problem of your choice, propose a solution using appropriate techniques, and implement a working prototype. This is a chance to be creative, solve real-world problems, and build a portfolio project.

2. Objectives

• Demonstrate understanding of image processing and computer vision principles
• Design and implement a complete vision pipeline (e.g., classification, detection, segmentation)
• Practice software engineering and project organization
• Evaluate your results and reflect on model limitations and performance

3. Suggested Topics

• Document layout analysis using OpenCV and image classification
• Face mask detection system using YOLOv5 or SSD
• Medical image classification (e.g., pneumonia, skin lesions)
• Object tracking in video sequences (e.g., using optical flow)
• Handwritten digit/character recognition using CNNs
• Vehicle or pedestrian detection for smart traffic monitoring
• QR code or barcode reader system with preprocessing and decoding

4. Project Phases

4.1 Proposal (Week 11 - Day 1–3)

• Submit a 1-page project proposal (template provided)
• Include: Title, abstract, problem description, proposed method, expected outcomes
• Optional: Submit a short pitch video (1–2 mins)

4.2 Implementation (Week 11–12)

• Use Python + OpenCV or PyTorch/TensorFlow
• Organize your project folder: `/src`, `/data`, `/models`, `/notebooks`, `/results`
• Write clean, well-commented code
• Use GitHub or Google Colab for development and sharing

4.3 Evaluation (Week 12)

• Evaluate your system using appropriate metrics (accuracy, IoU, mAP, etc.)
• Visualize results clearly with plots, overlays, and error analysis
• Discuss limitations and potential improvements

4.4 Presentation

• Prepare a 5–8 minute presentation with slides (recorded or live)
• Summarize motivation, dataset, pipeline, results, and reflections
• Include code demos or visual walkthroughs

5. Deliverables

6. Grading Rubric

7. Tips for Success

• Start early and test your code incrementally
• Document all experiments and keep backups
• Use version control (e.g., GitHub)
• Be honest about what worked and what didn’t