Mansouri Abdelkhalek

Led the integration of Tattile smart cameras into access control systems, defining communication protocols. Developed C-based APIs and middleware to abstract camera hardware, enabling seamless integration with the core application.

Developed and maintained the T&S "Remote Work Station" project. Engineered custom Yocto build systems and integrated OTA updates for embedded Linux deployments.

Developed, integrated, and maintained real-time radar processing chains. Optimized C++ code for embedded targets to achieve real-time performance and built Yocto-based Linux OS for radar systems.

Designed GPU-accelerated parallel algorithms for NP-hard optimization problems, applying them to computer vision tasks like image segmentation and stereo matching. Published research in international conferences and journals.

Studied and analyzed Simultaneous Localization and Mapping (SLAM) algorithms based on Kalman filter. Implemented feature detection algorithms on Zynq-7000 and deployed VHDL code on the Zynq-7000 FPGA board.

Designed and developed an IP PBX system using Asterisk software. Created a web administration interface for user and feature management, and deployed telephony solutions on CentOS servers.

Completed a Ph.D. in Computer Science, with a thesis focused on generic heuristics on GPU for superpixel segmentation and application to optical flow estimation. This research involved advanced topics in computer vision and parallel computing.

Obtained a Master's degree with a thesis on object detection and implementation of SLAM on a reconfigurable FPGA, embedded in a SWARM of drones. The curriculum covered software development for embedded systems and computer vision.

Earned a Bachelor of Science in Computer Science Engineering. The engineering thesis involved the design and development of an IP PBX solution based on open-source software.