冷凝器清洗机器人的奇Gabor边缘检测算法研究  

Odd Gabor edge detection algorithm for condenser cleaning robot

作　　者：张莹[1,2] 王耀南[1] 

机构地区：[1]湖南大学电气与信息工程学院,长沙410082 [2]湘潭大学信息工程学院,湘潭411105

出　　处：《仪器仪表学报》2010年第7期1564-1568,共5页Chinese Journal of Scientific Instrument

基　　金：国家自然科学基金(60775047);国家自然科学基金重点项目(60835004);国家863计划(2007AA04Z244;2008AA04Z214)资助

摘　　要：冷凝器管口的边缘检测是实现清洗机器人喷枪口精确定位的关键步骤。由于二维Gabor函数是能够取得空域和频域联合测不准原理下限的唯一函数,本文提出了二维奇Gabor算法应用于大型冷凝器清洗机器人清洗中的边缘检测,它克服了傅里叶变换对信号的局部特征没有分析能力的缺陷,可以在多个尺度上对图像的边缘进行检测。文章对检测的性能做了详细分析,实验结果证明该方法具有良好的边缘检测效果,其性能优于Sobel、Prewitt、Roberts、Log等传统边缘检测算法。Condenser nozzle edge detection is the key step for Because 2D Gabor function is the only function that is able to and frequency domains. In this paper, we propose an 2D Odd condenser cleaning robot to locate all tubes correctly. achieve the minimum Gabor edge detection uncertainty principle in spatial algorithm for the auto-cleaning robot of large condensers. The proposed algorithm overcomes the defects of Fourier transform, i.e. lacking of capa- bility to analyze the local features of the signal, and carries out edge detection in multi-scales. Then detailed analysis to 2D Odd Gabor performance is carried out and experimental results show its performance is better than Sobel, Pre-witt, Roberts, Log and other traditional algorithms in edge detection.

关 键 词：冷凝器 边缘检测 二维奇Gabor变换 清洗机器人 视觉定位 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置]