基于图像分割技术的皮革面积检测算法

Leather Area Detection Algorithm Based on Image Segmentation Technology

为解决目前皮革绷板干燥工序中皮革有效面积难以用传统方法检测的问题,本文提出了一种皮革轮廓分割方法,并设计皮革面积检测算法。首先,本文通过搭建试验平台模拟绷展状态下的羊革,利用单目相机获得羊革原图像,经过降噪、二值化完成图像的预处理。然后,利用腐蚀、膨胀处理实现对羊革轮廓与绷板夹的分割。最终,识别分割后羊革轮廓,进行羊革有效面积计算。试验结果表明,本算法可有效分割绷展状态下的羊革与绷板夹,得到羊革有效面积,基本满足自动化皮革绷板机的工艺需求。本算法不仅可进一步提高皮革得革率,增加制革企业经济效益,也为智能化皮革绷板机的设计提供了技术基础。

To solve the problem of limited leather area that is difficult to detect using traditional methods in the cur⁃rent leather stretch drying process,this paper proposed a leather contour segmentation method and designed a leather area detection algorithm.Firstly,this paper simulated the stretched state of sheep leather by building an experimental platform,and obtained the original image of the sheep leather using a monocular camera.The image was preprocessed through denoising and binarization.Then,using corrosion and expansion treatment to achieve segmentation of the sheep leather contour and the tension plate clamp.Finally,recognize the contour of the segmented sheep leather and calculate the effective area of the sheep leather.The experimental results show that this algorithm can effectively seg⁃ment the sheep leather and the tension plate clamp in a stretched state,obtain the sheep leather area,and basically meet the process requirements of automated leather tension plate machines.The algorithm can not only further improve the leather yield and increase the economic benefits of leather manufacturing enterprises,but also provide a theoretical basis for the research of intelligent leather stretcher machines.