ColourBlindAssist

This project implements a Color Blindness Assistance System that performs both Color Vision Deficiency (CVD) simulation and daltonization correction using Machado-style matrices and perceptual color space modeling. The system enhances color distinguishability for users with color blindness by transforming images and live video feeds in real time. It supports both static image processing and webcam-based correction pipelines. The implementation combines two complementary approaches. First, a Machado et al. matrix-based simulation model is used to reproduce Deuteranopia perception by transforming normalized RGB color space through physiologically motivated matrices. Second, a perceptual modeling pipeline using the Colorspacious library simulates varying levels of CVD severity and applies adaptive daltonization based on error redistribution. Daltonization is performed by computing the perceptual color error between the original image and its simulated CVD version, then redistributing that error back into the visible spectrum. This improves contrast between confusing colors while preserving overall visual structure. A tunable correction factor controls enhancement strength. The system includes an interactive real-time webcam mode where users can dynamically adjust severity levels using keyboard controls. Frame-by-frame processing converts input streams into corrected output with on-screen severity indicators. An image analysis mode generates comparison panels showing original, simulated, and daltonized outputs across multiple severity levels for evaluation and visualization. Core technologies used include OpenCV for image/video processing, NumPy for matrix operations, Matplotlib for visualization, and Colorspacious for perceptual color transforms. The design emphasizes modular processing pipelines so simulation and correction strategies can be swapped or extended. This project is intended as an assistive accessibility tool and a research-oriented prototype. Potential applications include AR accessibility filters, UI contrast enhancement layers, educational visualization tools, and wearable vision systems.