基于改进生成对抗网络的铝板缺陷电涡流检测图像分割  

Aluminum Plate Defect Image Segmentation Using Improved Generative Adversarial Networks for Eddy Current Detection

作　　者：张琦 叶波[1,2] 罗思琦 曹弘贵[1,2] Zhang Qi;Ye Bo;Luo Siqi;Cao Honggui(Faculty of Information Engineering and Automation,Kunming University of Scienceand Technology,Kunming,Yunnan 650500,China;Yunnan Key Laboratory of Artificial Intelligence,Kunming University of Science and Technology,Kunming,Yunnan 650500,China)

机构地区：[1]昆明理工大学信息工程与自动化学院,云南昆明650500 [2]昆明理工大学云南省人工智能重点实验室,云南昆明650500

出　　处：《激光与光电子学进展》2021年第8期329-337,共9页Laser & Optoelectronics Progress

基　　金：云南省应用基础研究计划(2019FB081)。

摘　　要：针对铝板缺陷电涡流检测图像边缘区域不容易识别,同时检测图像存在背景噪声干扰的问题,提出了一种基于改进生成对抗网络的铝板缺陷电涡流检测图像分割方法。该方法建立在生成对抗网络图像分割模型的基础上,生成器部分采用U-Net模型的思想,在高低层特征融合之前采用注意力模块,调整低层特征与高层特征拼接时的权重,有利于提高图像特征信息的利用,增强目标特征并抑制背景特征。判别器部分用于区分网络生成的结果和人工标注的真实结果。所提方法采用Precision、Recall和F1作为评价指标。实验结果表明,与传统的图像分割方法相比,所提方法对铝板缺陷电涡流检测图像取得了较好的分割效果。To address the difficulty associated with identifying the edge area in aluminum plate defect eddy current inspection images,in which background noise is typically problematic,an image segmentation method for aluminum plate defect eddy current detection based on improved generative adversarial network is proposed.The proposed method is based on the generative adversarial network image segmentation model.The generator partly adopts the idea of the U-Net model.Prior to the fusion of high-and low-level features,an attention module is used to adjust the weight of both low-and high-level features.This weight adjustment improves the utilization of image feature information,enhances the target features,and suppresses background features.The discriminator network is used to distinguish the results generated by the network and actual manually labeled results.The proposed method uses Precision,Recall,and F1 as evaluation indicators.Compared with the traditional image segmentation methods,the proposed method achieves a better segmentation effect for aluminum plate defect eddy current inspection images.

关 键 词：机器视觉 电涡流检测 生成对抗网络 图像分割 注意力机制 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]