Tracking Object for the Mine Safety Monitoring System
Because the safety monitoring system is very useful to safely mining, it has received a considerable amount of attention. It is most important to observe the mining scene. So a mobile robot with a PTZ (Pan-Tilt-Zoom) camera is adopted. But there exists the vibration influence of the mobile robot, image stabilization becomes poor. In fact, human eyes can perfectly track an object from running vehicle. Therefore, this paper presents the modeling and implementation of tracking an object from a mobile robot based on biomimetic oculomotor control of human eye. Firstly, after anatomic neural loop and physiological mechanism of eye movement control are analyzed, the biomimetic oculomotor control model is obtained. The model needs two inputs. One is the robot attitude information detected by gyros. The other one is equal to the retinal slip obtained by image processing unit. Secondly, considering the particularities of tracking system, foreground model, segmentation of motion based on multiple thresholds, feature extraction of area and center of gravity of a motion region, and location of the tracked object are used to estimate the retinal slip of biomimetic model. Finally, an efficient tracking system based on biomimetic oculomotor control model is developed. In order to demonstrate the object tracking efficiency, the tracking experiments are carried out. The tracking results are satisfactory. It is a new effective method for the mine safety monitoring system.
mine safety monitoring tracking object mobile robot biomimetic oculomotor control of human eye
XIE Shaorong LUO Jun GONG Zhenbang
School of Mechatronics Engineering and Automation, Shanghai University, Shanghai 200072, China
国际会议
The 2007 International Symposium on Safety Science and Technology(2007采矿科学与安全技术国际学术会议)
河南焦作
英文
2153-2159
2007-04-17（万方平台首次上网日期，不代表论文的发表时间）