基于二次阈值分割的方向倒脚匹配工件识别  

Work-piece recognition of orientation chamfer matching based on twin-threshold segmentation

作　　者：李博[1] 陈燕[1] 邹湘军[1] 彭红星[1,2] 谈建豪[1] 蒋志林[1] 

机构地区：[1]华南农业大学工程学院南方农业机械与装备关键技术教育部重点实验室,广东广州510642 [2]华南农业大学信息学院,广东广州510642

出　　处：《计算机工程与设计》2016年第3期720-724,共5页Computer Engineering and Design

基　　金：广东省教育部产学研结合基金项目(2012B091000167);国家自然科学基金项目(31171457);广东省科技计划基金项目(2014A020208091)

摘　　要：为提高机器人在复杂背景下利用模板匹配对工件的识别率,提出基于二次阈值分割的方向倒脚匹配(orientation chamfer matching)方法。利用大津法对复杂背景的目标图像进行第一次图像阈值分割,使用迭代阈值法对目标图像进行第二次图像阈值分割,运用方向倒脚匹配的方法对工件目标进行识别。对选取的40张复杂背景下的工件图片进行测试与分析,工件的平均识别率达到72.75%,误检测降低到0.75%。实验结果表明,该改进算法与传统方向倒脚匹配相比性能有较大提升,在复杂背景下对工件识别有效可行。To enhance the robots identification rate of work-piece in the complex background,an algorithm of orientation chamfer matching based on the twin-threshold segmentation was proposed.First segmentation was carried out to process the target image in the complex background using Otsu method.And second image segmentation was carried out to process the target image using iterative threshold method with the purpose of segmenting the targets form the complex background.Orientation chamfer matching was used to recognize the targets.40 work-piece images were used to test this algorithm,the work-piece＇s average recognition rate reached 72.75% and the false detection reduced to 0.75%.Results of test show that method proposed is better than the traditional algorithm,and it is feasible and effective on recognizing work-pieces in the complex background.

关 键 词：复杂背景 方向倒脚匹配 二次阈值分割 工件识别 机器人 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]