Stereo Depth Estimation — 3D Vision / Spatial Computing

Built a stereo depth estimation system using a calibrated and rectified dual-camera rig with hardware-synced capture — ensuring both frames are captured at the exact same moment, eliminating motion artefacts that corrupt depth quality. The pipeline covers the full stereo vision stack: intrinsic and extrinsic camera calibration, stereo rectification to align epipolar geometry, and dense disparity computation to produce accurate per-pixel depth maps. The result is a reliable, real-world depth sensing solution without the cost and power constraints of active sensors like LiDAR or structured light — suitable for robotics, industrial inspection, autonomous navigation, and spatial computing applications.