摘要
根据灰度矩算子在目标成像前后的矩不变特性,利用Tabatabai等人提出的前三阶灰度矩建立了亚像素边缘检测算法的完备数学模型。通过实验对该算法的有效性和检测精度进行了研究,给出了工件的实测尺寸对比结果。实验表明:基于灰度矩算子的亚像素边缘检测算法比传统算子具有更高的检测定位精度,可满足图像目标高精度实时在线测量的要求。
The gray moment operator's invariant characteristics for target imaging fore and after are introduced. According to this characteristic and Tabatabai's first three sample gray moments, a closed mathematical model of subpixel edge detection operator is founded. An application of the proposed edge operator as an edge detector is provided, the validity and detecting precision are studied and the comparison results of parts are also given. The experiments show that the gray moment operator has higher detecting precision than traditional operators and the proposed method is very effective for real-time online measurements.
出处
《光学技术》
CAS
CSCD
2004年第6期693-695,698,共4页
Optical Technique
基金
上海市科学技术委员会重点科技资助项目(021111125)
关键词
灰度矩
亚像素
边缘检测
阶跃边缘
归一化
gray moment
subpixel
edge detection
step edge
normalization