基于拉普拉斯算法和梯度边缘特征的鱼货去鳞机表面磨损检测方法  

Surface Wear Detection Method for Fish Descaling Machine Based on Laplace Algorithm and Gradient Edge Features

作　　者：梁秀娟[1] 嵇海旭[1] LIANG Xiujuan;JI Haixu(School of Mechanical Engineering,Guangdong Ocean University,Zhanjiang 524008,China)

机构地区：[1]广东海洋大学机械工程学院,湛江524008

出　　处：《自动化与仪表》2024年第11期74-78,共5页Automation & Instrumentation

基　　金：广东海洋大学校级教育教学改革项目(PX-972023225)。

摘　　要：为了确保鱼货去鳞机的正常使用,需要检测其表面磨损情况。提出基于拉普拉斯算法和梯度边缘特征的鱼货去鳞机表面磨损检测方法。首先,采用布谷鸟搜索算法优化二维高斯函数,结合拉普拉斯算法与优化后的二维高斯函数对鱼货去鳞机表面图像展开卷积处理,以此增强磨损区域边缘细节信息,实现磨损区域边缘检测;其次,对磨损区域展开Gamma校正,在此基础上提取磨损区域的HOG特征与LSS特征,结合上述特征构成梯度边缘特征;最后,基于S-SVM建立鱼货去鳞机表面磨损检测模型,将梯度边缘特征输入模型内,实现鱼货去鳞机表面磨损检测。实验结果表明,所提方法可准确确定磨损边缘,高精度地完成磨损特征提取与检测。To ensure the normal use of the fish descaling machine,it is necessary to inspect its surface wear.A surface wear detection method for fish descaling machine based on Laplace algorithm and gradient edge features is proposed.Firstly,the cuckoo search algorithm is used to optimize the two-dimensional Gaussian function.Combining Laplace algorithm with the optimized two-dimensional Gaussian function,the surface image of the fish descaling machine is convolved to enhance the detailed information of the wear area edge and achieve edge detection of the wear area;Secondly,perform Gamma correction on the worn area,extract the HOG and LSS features of the worn area,and combine them to form gradient edge features;Finally,a surface wear detection model for the fish descaling machine is established based on S-SVM,and gradient edge features are input into the model to achieve surface wear detection of the fish descaling machine.The experimental results show that the proposed method can accurately determine the wear edge and achieve high-precision wear feature extraction and detection.

关 键 词：拉普拉斯算法 布谷鸟搜索算法 梯度边缘特征 鱼货去鳞机 磨损检测 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置]