摘要
针对网球收集机器人工作中识别准确率低下问题,提出一种基于网球颜色和轮廓特征的识别算法。通过分析网球图像的颜色空间,得到网球与背景的二值图像,结合网球区域特征做进一步精准识别。在重叠网球识别上,提出区域分割和轮廓拟合的算法,对重叠区域提取分割、拟合还原网球轮廓。通过实验验证算法可行性,结果表明该算法可在不同环境下精确识别网球。
Aiming at the low recognition accuracy of tennis collection robot,a tennis recognition algorithm based on tennis color and contour features is proposed.By analyzing the color space of tennis image,the binary image of tennis and background is obtained.The binary image is optimized by morphological processing,and the tennis area features are combined for further accurate tennis recognition.In the recognition of overlapping tennis balls,an algorithm of region segmentation and contour fitting is proposed to extract,segment,fit and restore the contour of tennis balls.Realize the recognition of overlapping tennis balls.Finally,the feasibility of the algorithm is verified by experiments.The results show that the recognition algorithm can accurately recognize tennis indoors and outdoors.
作者
孙金风
申言鑫
杨智勇
陈龙
SUN Jinfeng;SHEN Yanxin;YANG Zhiyong;CHEN Long(School of Mechanical Engineering,Hubei Univ.of Tech.,Wuhan 430068,China)
出处
《湖北工业大学学报》
2023年第2期22-26,共5页
Journal of Hubei University of Technology
基金
国家自然科学基金(51907055)
湖北省教育厅中青年人才项目(Q20191404)。
关键词
网球识别
二值图像
形态学
区域分割
tennis recognition
binary image
morphology
region segmentation
