基于序贯相似度的AGV图像配准方法

AGV's Image Registration Algorithm Based on SSDA

在视觉引导AGV技术中,符号识别是关键。为实现AGV对地面符号快速、准确的识别,当AGV视觉系统获得的地面符号图像时,首先利用统计特征区域的黑点数来筛选模板,然后采用基于模板匹配的图像配准算法对筛选后的模板进行匹配,完成识别。经过仿真实验证明,该算法的匹配率普遍达到90%以上,基本满足了AGV模板识别、定位的需要,并且计算量较小、计算时间较短。

Symbol recognition is the key in visual-guided AGV. In order to achieve AGV recognition the symbol faster and accurate, when the AGV vision system to obtain images of the ground symbol, filters the model by use of counting the black point of character area at first, then adopt a new Image Registration Algorithm it＇ s based on image matching to match the template. These methods have been confirmed in the simulation, and the result shows that the match rate is above 90%. It satisfies the AGV＇ s model recognition, localization basically and efficiently.