西北工业大学 航海学院,西安,710072
[ "张晓阳, 硕士研究生, zhangxiaoyang" ]
[ "何军红, 副教授, 硕士生导师," ]
纸质出版:2025
移动端阅览
张晓阳, 何军红, 牛云, 等. 改进Canny算子在零件尺寸测量中的应用[J]. 机械科学与技术, 2025,44(4):625-631.
张晓阳, 何军红, 牛云, et al. Application of Improved Canny Operator in Size Measurement of Parts[J]. Mechanical Science and Technology for Aerospace Engineering, 2025, 44(4): 625-631.
张晓阳, 何军红, 牛云, 等. 改进Canny算子在零件尺寸测量中的应用[J]. 机械科学与技术, 2025,44(4):625-631. DOI: 10.13433/j.cnki.1003-8728.20230196.
张晓阳, 何军红, 牛云, et al. Application of Improved Canny Operator in Size Measurement of Parts[J]. Mechanical Science and Technology for Aerospace Engineering, 2025, 44(4): 625-631. DOI: 10.13433/j.cnki.1003-8728.20230196.
零件尺寸视觉测量目前多采用像素级的边缘检测算法
不能满足高精度的测量需求。为提高零件尺寸测量精度
提出一种基于改进Canny算子的亚像素级零件尺寸测量方法。首先采用自适应中值-高斯滤波对零件图像进行滤波处理
在保留较多边缘信息的同时有效去除噪声; 然后对图像进行梯度计算并通过二次插值寻找边缘点; 其次通过编写的链接算法对边缘点进行链接并进行双阈值处理
提取到图像亚像素边缘轮廓; 最后获取轮廓的最小外接矩形并得到零件尺寸。实验表明
该方法有效解决了零件轮廓的精确定位问题
测量误差在1个像素尺寸以内
在保证测量精度和测量效率的同时具有较高的稳定性。
At present
pixel level edge detection algorithm is often used in visual measurement of parts size
which cannot meet the high precision measurement requirements. To improve the precision of part size measurement
a subpixel level part size measurement method with the help of the improved Canny operator was proposed. Firstly
adaptive median-Gaussian filtering is used to filter the image of parts
which can effectively remove noise while preserving more edge information. Then the gradient of the image is calculated and the edge points are found by using the quadratic interpolation. Secondly
the link algorithm is used to link the edge points and perform double threshold processing to extract the image subpixel edge contour. Finally
the minimum external rectangle of the contour is obtained and the part size is obtained. Experimental results show that the present method can effectively solve the accurate positioning of the contour of parts
and the measurement error is below 1 pixel. It has high stability while ensuring the measurement accuracy and efficiency.
0
浏览量
0
下载量
0
CSCD
关联资源
相关文章
相关作者
相关机构
京公网安备11010602201714号