A visual sensing system integrating various sensory information

点击翻译按钮获取中文摘要

1. A new method for estimating the magnitude of the angular velocity of self-motion by measuring the average pressure on both inside and outside walls of a fluid-filled circular canaj was proposed. Our theory assuming the hydrostatic condition was discussed and proved by numerical simulations and experimental results.2. It is desirable for intelligent image sensing system to be able to control its gaze direction with regard to the motion of the target object. To realize the gaze control, we formulated a control criterion similar to that of human eye movements : saccade and smooth pursuit. Experimental results showed the effectiveness of our gaze control method. In these experiments, our system showed an important property of smart sensing system, i. e. the ability to dynamically select the optimum operating mode for the purpose of measuring.3. We proposed and demonstrated a newly extended real-time and flexible intelligent visual sensing system which was able to integrate visual sensory information with angular velocity information, and control camera gaze direction by both electronic and mechanical means simultaneously. The angular velocity sensor was composed of three vibratory gyroscopes ahad was able to detect three components of angular velocity. Experimental results showed that the system performs good fixation and moving object detection under gyro stabilized condition. It is desirable that the future intelligent visual sensing system mounted on locomotive machines can not only compensate the turbulence due to their movement, but also utilize the turbulence actively as the necessary information to recognize the outer world. As the results, such advanced signal processing will become more important technique.