摘要
在图像准确检测问题的研究中,由于在高精密仪器中,各种损伤极其细微,传统的利用模具高斯曲线拟合检测技术在进行这种破损检测时,往往会因为破损部位过小,造成特征混淆,特征寻优过程陷入不收敛的境地,从而降低了精密模具破损检测的准确性。为了避免上述缺陷,提出了一种边缘像素定位的精密模具破损检测算法。利用像素灰度差值方法,提取精密模具图像的像素特征,并将其作为精密模具破损检测的数据基础。利用一种边缘像素定位方法,计算精密模具破损区域边缘的空间位置,从而实现精密模具的破损检测。实验结果表明,利用改进算法进行精密模具破损检测,能够提高检测的精度,从而满足精密模具生产过程中质量检测的需求。
Accurate recognition of image was researched.This paper put forward a precision mold damage detection algorithm based on edge pixel location.Pixel gray level difference method was employed to extract precision mold image pixel features which were served as detection data base.A method of edge pixels location was used to calculate the spatial position of edge and realize precision mold damage detection.The experiment results show that the algorithm can improve the accuracy of identification.
出处
《计算机仿真》
CSCD
北大核心
2013年第8期237-240,共4页
Computer Simulation
关键词
计算机视觉
边缘像素
破损检测
精密模具
Computer vision
The edge pixels
Damage detection
Precision mould
