This project implements two fundamental approaches for computing disparity maps from stereo image pairs: Block Matching and Dynamic Programming. These methods are used to estimate depth information from stereo images, a core problem in computer vision.

A Matlab implementation of Block Matching Algorithm for stereo matching. It uses matrix operations to make the code faster.

Low-Rank Regularized Image Denoising

A C++ implementation of Block Matching Algorithm for stereo matching. It uses the "SAD" (Sum of Absolute Differences) similarity metric.

A C++ implementation of Block Matching Algorithm for image denoising. It uses the "SAD" (Sum of Absolute Differences) similarity metric.

Numba based GPU block matching

Motion visualization with macroblock matching using OpenCV in python

YAVC (Yet Another Video Compressor) is a video compressor, that compresses raw frames

An implementation of the block matching stereo vision algorithm.

depth map computation

3 Step Search H.264 Compression Block Matching Algorithm. The only Python implementation on the public internet.

Heightmap reconstruction from stereo pair/anaglyph (proof of concept)

Deep learning based image denoising using tensorflow/Keras combined with block matching

Object tracking with OpenCV based on stereo camera images