摘要
为实现零件图像的边缘检测,针对传统基于微分的边缘检测算法存在边缘点定位不准确、角点漏检等不足,提出一种基于果蝇优化算法(fruit fly optimization algorithm,FOA)的零件图像边缘检测算法。该算法首先通过Canny算子得到边缘点的先验知识,再利用希尔伯特变换提取角点信息,以边缘点和角点信息作为启发信息,建立基于FOA的零件图像边缘检测模型,最后通过随机平均移动机制和循环终止条件得到图像的单像素边缘。经实验验证,算法在无噪声边缘检测的条件下,相比传统的Canny算子,在零件图像检测的精度和准确性上有较大提升,可应用于工业零件的高精度无损检测。
To realize edge detection for images of mechanical parts,aiming at the shortcomings of the traditional differential-based edge detection algorithms,such as inaccurate locating edge points and missing corner points,an edge detection method based on fruit fly optimization algorithm(FOA)was proposed.This method first uses the Canny edge detection operator to obtain the prior knowledge of the edge points and uses the Hilbert transform to extract the corners as heuristic information.The edge detection algorithm is based on the fruit fly optimization algorithm.Finally,through the random average movement mechanism,the loop terminates the query conditions at the optimal value and the single-pixel edge of the image.It has been verified by experiments that under the condition of noise-free edge detection,compared with the traditional Canny edge detection,the algorithm has a great improvement in the precision and accuracy of image detection of parts,and can be applied to high-precision non-destructive detection of industrial parts.
作者
谢昕
王伟如
万天成
江勋绎
胡锋平
XIE Xin;WANG Wei-ru;WAN Tian-cheng;JIANG Xun-yi;HU Feng-ping(School of Information Engineering,East China Jiaotong University,Nanchang 330013,China;School of Civil Engineering,East China Jiaotong University,Nanchang 330013,China)
出处
《科学技术与工程》
北大核心
2021年第5期1948-1956,共9页
Science Technology and Engineering
基金
国家自然科学基金(61762037,61872141)
江西省自然科学基金(20181BAB206037)。
