Abstract: To improve the initivative and stability of Pan-Tilt-Zoom(PTZ) adjustment in visual system under constant illumination condition, Active Pan-Tilt-Zoom(APTZ) adjustment method based on Local particle Filtering Pre-Location(LFPL) under constant illumination condition is proposed. Target pre-location mehtod based on local particle filter is used to realize automatic estimation calibratiing for moving target. Initivative of PTZ adjusting is improved and nonlinear tracking problem can be settled. Dynamic selecting illumination-inviarants as filtering particles can overcome the affect of illumination varying and noise in target pre-location algorithm and robust of visual system is improved. Fuzzy control method is used to control pan and tilt to improve the stability of visual tracking system. The correctness and accuracy are tested by experiments. Long distance tracking for varying speed moving target using this method shows it is more stabile than the traditional method. Moving target racking under varying illumination and noise conditions shows ideal result.

Key words: Active Pan-Tilt-Zoom(APTZ) adjustment, Local particle Filtering Pre-Location(LFPL), sampling, Fuzzy control

摘要： 为提高光照恒定情况下视觉系统中PTZ调节的主动性和稳定性，提出恒定光照下基于LFPL的APTZ调节方法。采用基于局部粒子滤波的目标预定位方法，对运动目标实现自动估计的标定，提高系统调节的主动性，解决非线性跟踪问题。动态选取光照不变特征滤波粒子克服了光照变化和噪声等因素对目标预定位方法的影响，增强视觉系统的鲁棒性。对水平角和抑角采用Fuzzy控制方法，提高视觉跟踪系统的稳定性。实验结果表明，该方法是正确有效的，使用该系统对变速运动目标的长距离跟踪结果较传统方法更稳定，在光照变化和噪声条件下的运动目标跟踪实验也取得较好的结果。

关键词: 主动PTZ调节, 局部粒子滤波预定位, 采样, Fuzzy控制

CLC Number: 

• TP301