基于参数可调均值迁移滤波实现筒子纱的识别检测  

Recognition and Detection of Cheese Based on Edge Preserving Filtering Algorithm

作　　者：李靖宇 沈丹峰[1] 王玉 李耀杰 LI Jingyu;SHEN Danfeng;WANG Yu;LI Yaojie(School of Mechanical and Electrical Engineering,Xi′an Polytechnic University,Xi′an 710600,China)

机构地区：[1]西安工程大学机电工程学院,陕西西安710600

出　　处：《轻工机械》2021年第3期63-68,72,共7页Light Industry Machinery

基　　金：陕西省科技厅重点研发计划(2018TSCXL-GY-03-02)。

摘　　要：为了提高机器视觉对纱筒图像识别检测的准确率,课题组提出了一种参数可调均值迁移滤波算法。在图像采样过程中,先对源图像进行锐化处理,再对图像空间窗口与颜色窗口同时进行卷积操作,筒子纱的边缘轮廓得到完整保留,且通过滤波处理去除图像噪声;采用Canny算子边缘检测,完成对筒子纱的轮廓提取;几何矩计算轮廓质心和横纵比过滤判别筒子纱的放置状态。实验结果表明:该方法在图像处理过程中准确、可靠,能够满足图像识别检测的要求。In order to improve the accuracy of machine vision on screen image recognition, the research group proposed a parameter-adjustable mean shift filtering algorithm. In the process of image sampling, first the source image was sharpened, then the image space window and color window were convolved simultaneously. The edge contour of cheese was completely preserved, and the image noise was removed by filtering. Canny operator edge detection was used to complete the contour extraction of cheese;geometric moment calculation of contour centroid and transverse/longitudinal ratio filtering were used to judge the placement state of cheese. The experimental results show that this method is accurate and reliable in image processing, and can meet the requirements of image recognition and detection.

关 键 词：机器视觉 参数可调均值迁移滤波算法 锐化处理 边缘检测 几何矩 横纵比过滤 

分 类 号：TP391[自动化与计算机技术—计算机应用技术] TH162[自动化与计算机技术—计算机科学与技术]