Rachid Deriche

Institut national de recherche en sciences et technologies du numérique, Centre Inria d'Université Côte d'Azur

Papers

Total Citations

445

H-Index

About

Rachid Deriche is a pioneering figure in computer vision, whose work has fundamentally shaped how machines perceive and navigate the world. His research centers on feature extraction, motion analysis, and 3D reconstruction, with a particular emphasis on robust algorithms for tracking and matching visual data. Deriche’s most influential contribution is his seminal work on "Tracking Line Segments" (1990), which has garnered over 350 citations and remains a cornerstone for dynamic scene analysis. This method provided a principled framework for following features across image sequences, enabling advances in autonomous navigation and robotics. He further extended these ideas to stereo vision, developing an efficient approach for matching high-curvature points using two or three cameras, a technique critical for mobile robot localization. Deriche also made notable strides in underwater navigation, where he introduced terrain-based matching algorithms using sonar bathymetric profiles to achieve precise absolute positioning. As a key contributor to the ESPRIT project 940, he helped design the Depth and Motion Analysis Machine, a hardware implementation that integrated edge detection, token tracking, and trinocular stereo. Through his rigorous, application-driven research, Deriche has left an indelible mark on how computers extract depth and motion from visual input, inspiring generations of researchers in autonomous systems.