Change Detection

Overview

Developed algorithms to detect scene changes across frames for surveillance and monitoring applications.

Responsibilities

• Designed frame differencing pipelines
• Handled illumination variations and noise
• Reduced false positives using post-processing

Approach

• Frame differencing
• Background modeling
• Morphological filtering

Pipeline

flowchart LR
    A[Frame t] --> C[Frame Differencing]
    B[Frame t+1] --> C
    C --> D[Thresholding]
    D --> E[Morphological Filtering]
    E --> F[Change Mask]

Tech

OpenCV · NumPy

Impact

• Enabled reliable detection of motion and scene changes
• Reduced noise-induced false detections
• Improved robustness under varying lighting conditions