基于计算机视觉的锻件尺寸检测  

Study on the Edge Detection of Forging Based on Computer Vision

作　　者：马俊福[1] 周晓军[1] 徐志农[1] 迟军[1] 

机构地区：[1]浙江大学机械与能源工程学院,杭州310027

出　　处：《组合机床与自动化加工技术》2007年第11期48-50,共3页Modular Machine Tool & Automatic Manufacturing Technique

摘　　要：利用计算机视觉系统获取了锻件的原始图像,根据白色背景下锻件原始图像r分量的灰度直方图确定了图像分割的阈值,并进行了锻件图像的分割。采用最大均匀性平滑法对分割后的锻件图进行了滤波,结果表明最大均匀性平滑法不会损伤锻件边缘形状。利用跟踪虫法对锻件边缘进行了跟踪,结果表明利用该锻件边缘跟踪虫法能提高锻件尺寸检测的精度。Computer Vision System is employed to acquire the original forging image, according to the r component of the histogram of the original forging image, the threshold is determined to segment the original forging image. The maximum homogeneity smoothing method is adopted to filter the segmented forging image, which has shown that the shape of the forging edge could not be damaged by this method. Tracking bug method is used to track the edge of the forging, which has shown that the precision of the forging size detection could be improved by using the edge tracking bug method.

关 键 词：计算机视觉 边缘 检测 尺寸 跟踪虫法 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]