瓷砖表面非常规尺寸瑕疵的检测  

Unconventional Size Defects Detection for Tile Surface

作　　者：陈凯 鲁涵统 程浩宇 方梦园 CHEN Kai;LU Han-Tong;CHENG Hao-Yu;FANG Meng-Yuan(School of Information Science and Technology,Zhejiang Sci-Tech University,Hangzhou 310018,China)

机构地区：[1]浙江理工大学信息学院,杭州310018

出　　处：《计算机系统应用》2022年第11期192-198,共7页Computer Systems & Applications

基　　金：国家自然科学基金(61903338);浙江省自然科学基金(LQ19F030015);辽宁省自然科学基金(2019-KF-23-02)

摘　　要：与普通目标检测任务不同,瓷砖表面瑕疵检测的困难之处在于检测小尺寸和大长宽比等非常规尺寸的目标.为了解决这两个问题,本文提出了一种基于改进Cascade R-CNN的新型瓷砖表面瑕疵检测算法.为了提高模型对小瑕疵的检测能力,本文模型利用侧向连接结构进行上下层语义信息的融合,使用可切换空洞率的空洞卷积来增加模型的感受野;为了提高模型对于大长宽比瑕疵的检测能力,本文模型在标准卷积上引入偏移域以更好提取目标特征信息.此外,本文模型调整Cascade R-CNN框架中预选锚框的大小和长宽比例.实验结果表明,在从瓷砖工厂收集的数据集上,本文所提出算法的平均精度均值(mean average precision,mAP)达到了73.5%,比改进前的Cascade RCNN模型提高了9.7%.本文实验代码可从以下链接获取:https://github.com/mashibin/Ceramic-tile-defect-detection.Different from ordinary object detection tasks,the difficulty of detecting tile surface defects lies in the detection of unconventional size objects,such as small-sized objects and objects with large aspect ratios.To solve these two problems,this study proposes a new type of tile surface defect detection algorithm based on improved Cascade R-CNN.To improve the detection ability for small defects,the model in this study uses the lateral connection structure to fuse the semantic information of the upper and lower layers and applies the dilated convolution with switchable dilation rates to increase the receptive field of the model.To improve the detection ability for defects with large aspect ratios,the proposed model introduces an offset field on the standard convolution to better extract the object feature information.In addition,the model adjusts the size and length of the pre-selected anchor box in the Cascade R-CNN framework.The experimental results show that on the dataset collected from the tile factory,the mean average precision(mAP)of the proposed algorithm reaches 73.5%,which is 9.7%higher than that of the Cascade R-CNN model before improvement.The experimental code of this study is available at:https://github.com/mashibin/Ceramic-tile-defect-detection.

关 键 词：目标检测 Cascade R-CNN 侧向连接 空洞卷积 偏移域 深度学习 卷积神经网络 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] TQ174.765[自动化与计算机技术—计算机科学与技术]