The session, held in the Smart Meeting Room at the FTUI Dean’s Building, focused on the theme “Advanced Imaging Technology for Optical Flow Analysis and Applications Based on Artificial Intelligence.” Dr. Hicham discussed the fundamentals of Computer Vision, including topics like Feature Detection and Tracking, Geometry in Computer Vision, and Mapping and 3D Reconstruction.
He explained that Feature Detection is the process of identifying significant points in an image, such as corners, edges, or unique patterns, which can be used for further analysis. Meanwhile, Feature Tracking involves monitoring the movement of detected features across frames in a video sequence.
Regarding geometry, Dr. Hicham outlined key concepts such as camera projection, calibration, and epipolar geometry. “These are crucial for transforming real-world 3D coordinates into 2D image representations. Techniques like triangulation estimate the 3D position of an object based on observations from two or more image views. Homography allows mapping points from one plane to another within an image, enabling transformations like rotation or translation,” he explained.
Mapping, he added, is the process of creating maps based on visual data, widely used in robotics to build environmental maps while tracking a robot’s position within it. 3D Reconstruction involves generating three-dimensional models of objects or environments using image data, typically employing techniques like stereo vision or multi-view reconstruction.
Dr. Hicham also shared his research, including:
- Visual Servoing: Using cameras to control a robot’s movements based on visual feedback. This research explores intelligent approaches, such as predicting object positions and adjusting paths to handle occlusions effectively.
- Multi-Robot Coordination and Task Sequencing: Developing algorithms to sequence tasks in multi-robot systems.
- Rotation Estimation Using Spherical Moments: Providing closed-form solutions for estimating 3D rotation using spherical moments.
- Motorcycle Attitude Estimation: Developing vision-based systems to estimate a motorcycle’s orientation, including lean and yaw angles, for applications like rider safety systems and vehicle dynamics analysis.
This guest lecture was part of the FMF agenda organized by AIDE RC. Attendees included FTUI postgraduate students, AIDE RC research centre members, and related faculty. The event served as a follow-up to promote international collaboration between FTUI and Université Évry Paris-Saclay.
In a separate statement, FTUI Dean Prof. Dr. Ir. Heri Hermansyah, S.T., M.Eng., IPU., remarked, “We hope this activity provides an avenue for students, particularly FTUI postgraduate students, to gain broader insights into AI, especially in computer vision applications.”
***
Public Communications Office
Faculty of Engineering, Universitas Indonesia
