Visual inspections of components require repetitive high-speed, highaccuracy checking to identify defects and surface imperfections that occur due to fluctuations in manufacturing processes. Visual inspection robots capable of learning and equipped with advanced intelligent sensors that outperform human eyes facilitate unmanned inspections, improved quality and precision, and fully automated assembly processes.

Visual inspection using a robot eye

Assembly robots require force sensors that detect applied pressures and motion-control technologies that reproduce the movements of human fingers. The development of coordination and feedback-control technologies for 3D visual information along with motion-planning technologies based on visual and pressure-tactile information is expected to lead to the creation of multifunctional and learning robots.

Face detection process and automatic face-tracking camera

Canon has developed its proprietary Face Detection Algorithm as a robot elemental technology. Local facial features are extracted and then hierarchically integrated, learned and recognized. By combining this high-speed face detection and tracking algorithm with coordinated software- and hardware-control technologies, Canon has developed an automatic face-tracking camera that achieves the world's highest level of response performance.