基于DSA-BCNN的毫米波图像隐藏物检测  

Concealed object detection from millimeter wave images based on DSA-BCNN

作　　者：胡川飞 王永雄[1] 李冬 高天天 HU Chuanfei;WANG Yongxiong;LI Dong;GAO Tiantian(School of Optical-Electrical and Computer Engineering,University of Shanghai for Science and Technology,Shanghai 200093,China;College of Arts and Sciences,Shanghai Polytechnic University,Shanghai 201209,China;Qingyi Its,Shanghai 201306,China)

机构地区：[1]上海理工大学光电信息与计算机工程学院,上海200093 [2]上海第二工业大学文理学部,上海201209 [3]擎翌(上海)智能科技有限公司,上海201306

出　　处：《光学技术》2021年第2期178-186,共9页Optical Technique

基　　金：国家自然科学基金面上项目(61673276)。

摘　　要：自动检测毫米波图像中被检人员是否携带隐藏物,是实现智能毫米波安检系统的重要技术之一。针对隐藏物在毫米波图像中的特征局部性和低辨识性问题,提出一种动态自注意力的双线性卷积神经网络,能够仅以图像级标签训练,实现图像中隐藏物的存在检测。引入自注意力机制以引导网络对隐藏物区域进行特征提取,增强网络刻画全局信息的能力;双线性池化构建的二阶特征丰富了网络对隐藏物与非检测区域的细微差异表征。实验结果验证了所提出方法对隐藏物检测的有效性,在各项评价指标上均高于其他基于卷积神经网络的方法,准确率达到93.6%。Detecting the concealed object from millimeter wave images is one of the key techniques to construct an intelligent millimeter wave based security inspection.To address the issue that the concealed objects are inspected hardly due to their locality and low identifiability in the millimeter wave images,a dynamic self-attentive bilinear convolutional neural network(DSA-BCNN)is proposed to train with image-level labels to detect the concealed objects.Self-attention mechanism is utilized to guide network to extract the features from concealed object regions,which enhances the network capability to depict the global information.Simultaneously,the second order features are constructed by bilinear pooling to enrich the representation of subtle differences between concealed objects and non-detected regions.Experimental results verify the propose method effectiveness,which is superior than others in terms of each evaluation metric,and the accuracy is 93.6%.

关 键 词：毫米波图像 隐藏物检测 卷积神经网络 自注意力机制 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]