当前位置: X-MOL 学术J. Visual Commun. Image Represent. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Image quality enhanced recognition of laser cavity based on improved random hough transform
Journal of Visual Communication and Image Representation ( IF 2.6 ) Pub Date : 2019-11-21 , DOI: 10.1016/j.jvcir.2019.102679
Yi-Bin He , Wen-Hao Xiong , Han-Xin Chen , Cheng-Hao Cao , Wengjian Huang , Liu Yang , Li Zeng , Qiao-Shen Dai , Yu-Chen Chen

The tedious measurement of arc parts with low accuracy is serious problem in the traditional industrial measurement. In this paper, the method for the image quality enhanced recognition with digital image processing technology is developed by changing the direct detection of arc parts of the laser cavity into the detection of contour curve in the image. A circular arc fitting algorithm based on the improved random hough transform (RHT) is proposed to improve the disadvantages of RHT such as strong noise disturbance, high requirements for the extraction of contour continuity and slow calculation speed. The differences between the distance from the center of the fitting circle to all points in the testing area and the fitting radius were calculated. The minimum value was obtained to determine the optimal fitting circular arc. The algorithm is tested and applied to detect the actual workpiece. It is demonstrated that the accuracy and speed of cavity detection are much better by comparision with the traditional algorithm by the proposed improved algorithm.



中文翻译:

基于改进的随机霍夫变换的图像质量增强了对激光腔的识别

在传统的工业测量中,低精度的弧形零件的繁琐测量是一个严重的问题。本文通过将激光腔的弧形部分的直接检测转换为图像轮廓曲线的检测,开发了一种利用数字图像处理技术进行图像质量增强识别的方法。提出了一种基于改进的随机霍夫变换(RHT)的圆弧拟合算法,以解决噪声干扰大,轮廓连续性提取要求高,计算速度慢等缺点。计算了从拟合圆心到测试区域所有点的距离与拟合半径之间的差。获得最小值以确定最佳拟合圆弧。测试该算法并将其应用于检测实际工件。通过与改进算法的比较,证明与传统算法相比,腔检测的精度和速度要好得多。

更新日期:2019-11-21
down
wechat
bug