基于颜色和纹理特征匹配的集装箱视觉识别与定位方法  

Container Visual Recognition and Positioning Method Based on Color and Texture Feature Matching

作　　者：闫兴达 冯云剑 

机构地区：[1]东南大学自动化学院,江苏南京210096

出　　处：《工业控制计算机》2024年第1期100-102,共3页Industrial Control Computer

摘　　要：集装箱自动化装卸是港口自动化中的关键任务,其中集装箱的视觉识别与定位是实现自动化装卸的关键技术之一。针对传统模板匹配算法在多尺度及光照变化情况下效果不佳的问题,将基于颜色特征块匹配算法与基于数学形态学纹理特征提取算法相结合,提出一种基于颜色和纹理特征匹配的集装箱视觉识别与定位算法。利用颜色块匹配的方法对集装箱进行初步定位,再根据集装箱表面凸起纹理特征进行精确定位。实验结果表明,该算法在满足实时性的基础上能有效实现集装箱的识别与定位,且对尺度及光照变化具有良好的鲁棒性。Automated loading and unloading of containers is a key task in port automation,and container recognition is one of the key technologies for achieving automated loading and unloading.To address the problem of poor performance of traditional template matching algorithms under multi-scale and lighting variation conditions,an algorithm based on color and texture feature matching is proposed,which combines a color-based feature block matching algorithm with a mathematical morphological texture feature extraction algorithm.The algorithm uses block matching to perform preliminary positioning of the container and uses a method to extract convex texture features on the container surface for precise positioning.The result shows that the algorithm can effectively recognize and locate containers while meeting real-time requirements and has good robustness to scale and lighting changes.

关 键 词：港口自动化 集装箱识别 块匹配 数学形态学 

分 类 号：U693.5[交通运输工程—港口、海岸及近海工程] TP391.41[交通运输工程—船舶与海洋工程]