Koichiro Deguchi
Papers
19
Total Citations
366
H-Index
9
About
Koichiro Deguchi is a pioneering researcher in computer vision and robotics, whose career has been defined by his groundbreaking contributions to visual servoing and vision-guided robot control. His work centers on enabling robots to interpret dynamic visual information directly to guide precise motion control, eliminating the need for complex 3D object models or explicit image feature correspondence — a significant practical advancement in the field. Deguchi's most influential contribution, "Optimal Motion Control for Image-Based Visual Servoing by Decoupling Translation and Rotation" (2002, 139 citations), introduced an elegant mathematical framework that dramatically improved the stability and efficiency of image-based control by separating translational and rotational camera movements. This insight addressed longstanding challenges in convergence and singularity that had limited earlier approaches. His earlier foundational work throughout the late 1990s established the eigenspace-based interpretation of dynamic images, creating robust theoretical underpinnings that the community continues to build upon. Beyond classical servoing, Deguchi extended his methods to sophisticated applications including simultaneous robot-object motion control, ball-catching robots employing just-in-time visual strategies, and biped humanoid robots using gaze-stabilization mechanisms inspired by human visual cognition. With over 300 cumulative citations across his portfolio, his research represents essential reading for anyone pursuing robotics, machine vision, or autonomous systems.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4A goal oriented just-in-time visual servoing for ball catching robot arm21 citations · 2008
- 5
- 6
- 7
- 8
- 92DOF motion stabilization of biped robot by gaze control strategy11 citations · 2005
- 10