摘要
在多目标图像分割中,经常会出现由于目标图像粘连度分布不均而影响图像分割精度的情况。通过研究标记分水岭算法,提出一种结合形态学运算和距离变换的改进的分水岭分割方法。该算法在处理粘连程度高的目标图像时,能够将粘连图像精确地分割出来。实验结果表明,将该算法应用到农产品的混合水果和医药领域药盒的无序分拣中,能够高效地解决粘连图像过分割的问题,且分割效率很高。
In multi-target image segmentation,the uneven distribution of adhesion degree of target image often affects the segmentation accuracy.By studying marking watershed algorithm,an improved watershed segmentation method based on morphological operation and distance transform is proposed.Using this algorithm,the target images with high adhesion could be accurately segmented.The experimental results show that when the algorithm is applied to the disordered sorting of mixed fruits of agricultural products and medicine boxes,it can solve the problem of over-segmentation of adhesion images efficiently.
作者
张文飞
韩建海
郭冰菁
李向攀
刘赛赛
Zhang Wenfei;Han Jianhai;Guo Bingjing;Li Xiangpan;Liu Saisai(School of Mechatronics Engineering,Henan University of Science and Technology,Luoyang 471003,Henan,China;Henan Provincial Key Laboratory of Robotics and Intelligent Systems,Luoyang 471003,Henan,China;Zhengzhou Institute of Science and Technology,Zhengzhou 450000,Henan,China)
出处
《计算机应用与软件》
北大核心
2021年第6期243-248,共6页
Computer Applications and Software
基金
河南省自然科学基金项目(162300410082)
河南省教育厅科学技术研究重点项目(14A460025)。
关键词
粘连度
形态学运算
距离变换
过分割
Adhesion degree
Morphological operation
Distance transformation
Over-segmentation
