Footprint Pressure Image Retrieval Algorithm Based on Multi-scale Self-attention Convolution
ZHU Ming1,2, WANG Tongsheng2, WANG Nian1, TANG Jun1,2, LU Xilong3
1. Key Laboratory of Intelligent Computing and Signal Processing,Ministry of Education,Anhui University,Hefei 230601;
2. School of Electronics and Information Engineering,Anhui University,Hefei 230601;
3. Institute of Forensic Science of China,Ministry of Public Security,Beijing 100038
To improve the accuracy of footprint pressure image retrieval,a footprint pressure image retrieval algorithm based on multi-scale self-attention convolution is proposed.Firstly,preprocessing operations,such as angle correction,alignment and erasure,are carried out to reduce the influence of image angle on feature extraction.Secondly,the discriminative features are extracted adaptively by the multi-scale self-attention convolution module composed of the hole convolution with multiple parallel branches and the self-attention module.Finally,the common incomplete scoring matrix is obtained by the incomplete scoring module composed of global feature branches and incomplete score mask branches.The discriminative features are weighted and combined via the common incomplete scoring matrix to improve the attention of the network to the common visible area of incomplete footprints.The experimental results show that the proposed algorithm produces higher first hit accuracy and average retrieval accuracy on the constructed FootPrintImage dataset compared with some existing image retrieval methods.
[1] ZHANG H,WANG X L,HE Z X,et al.Weighted Softmax Loss for Face Recognition via Cosine Distance//Proc of the Chinese Confe-rence on Biometric Recognition.Berlin,Germany:Springer,2018:340-348.
[2] ADAMOVI? S,MIšKOVIC V,MA?EK N,et al.An Efficient Novel Approach for Iris Recognition Based on Stylometric Features and Machine Learning Techniques.Future Generation Computer Systems,2020,107:144-157.
[3] DENG M Q,CHEN Q F.Human Gait Recognition Based on Deterministic Learning through Multiple Views Fusion.Pattern Recognition Letters,2016,78:56-63.
[4] KENNEDY R B.Uniqueness of Bare Feet and Its Use as a Possible Means of Identification.Forensic Science International,1996,82(1):81-87.
[5] BAO W X,WANG Y F,WANG N,et al.Optical Footprint Image Recognition Algorithm Based on Metric Learning and SVM//Proc of the International Conference on Computer Engineering and Application.Washington,USA:IEEE,2020:864-868.
[6] WANG X N,WANG H Y,CHENG Q,et al.Single 2D Pressure Footprint Based Person Identification//Proc of the IEEE International Joint Conference on Biometrics.Washington,USA:IEEE,2017:413-419.
[7] 张 浩.基于稀疏表示的足迹花纹图像检索算法研究.硕士学位论文.大连:大连海事大学,2017.
(ZHANG H.Research on Sparse Representation Based Shoeprint Retrieval Algorithm.Master Dissertation.Dalian,China:Dalian Maritime University,2017.)
[8] KHOKHER R,SINGH R C,KUMAR R.Footprint Recognition with Principal Component Analysis and Independent Component Analysis.Macromolecular Symposia,2015,347(1):16-26.
[9] 陶一宁,苏 峰,袁培江,等.联合足迹识别与监控视频分析的智能刑侦系统.北京航空航天大学学报,2020,46(9):1730-1738.
(TAO Y N,SU F,YUAN P J,et al.Intelligent Criminal Investigation System Based on Both Footprint Recognition and Surveillance Video Analysis.Journal of Beijing University of Aeronautics and Astronautics,2020,46(9):1730-1738.)
[10] 张 艳,王 乔,王 年,等.基于多模特征的足迹识别算法.华中科技大学学报(自然科学版),2019,47(5):73-78.
(ZHANG Y,WANG Q,WANG N,et al.Footprint Recognition Algorithm Based on Multi-modal Features.Huazhong University of Science and Technology(Natural Science Edition),2019,47(5):73-78.)
[11] 陈 扬,曾 诚,程 成,等.一种基于CNN的足迹图像检索与匹配方法.南京师范大学学报(工程技术版),2018,18(3):39-45.
(CHEN Y,ZENG C,CHENG C,et al.A CNN-Based Approach to Footprint Image Retrieval and Matching.Journal of Nanjing Normal University(Engineering and Technology Edition),2018,18(3):39-45.)
[12] 史力民,李 硕,赵悦岑.基于深度学习的赤足迹性别自动分析研究.中国刑警学院学报,2018(3):97-99.
(SHI L M,LI S,ZHAO Y C.Research on Automatic Analysis of Red Footprint Gender Based on Deep Learning.Journal of Criminal Investigation Police University of China,2018(3):97-99.)
[13] SZEGEDY C,LIU W,JIA Y Q,et al.Going Deeper with Convolutions//Proc of the IEEE Conference on Computer Vision and Pattern Recognition.Washington,USA:IEEE,2015.DOI:10.1109/CVPR.2015.7298594.
[14] HERMANS A,BEYER L,LEIBE B.In Defense of the Triplet Loss for Person Re-identification[C/OL].[2020-08-21].https://arxiv.org/pdf/1703.07737.pdf.
[15] KINGMA D P,BA J.Adam:A Method for Stochastic Optimization[C/OL].[2020-08-12].https://arxiv.org/abs/1412.6980.
[16] FAN X,LUO H,ZHANG X,et al.SCPNet:Spatial-Channel Pa-rallelism Network for Joint Holistic and Partial Person Re-identification//Proc of the Asian Conference on Computer Vision.Berlin,Germany:Springer,2018:19-34.
[17] DAI Z Z,CHEN M Q,GU X D,et al.Batch Feature Erasing for Person Re-identification and Beyond//Proc of the IEEE/CVF International Conference on Computer Vision.Washington,USA:IEEE,2019:3691-3701.
[18] WANG G S,YUAN Y F,CHEN X,et al.Learning Discriminative Features with Multiple Granularities for Person Re-identification//Proc of the 26th ACM International Conference on Multimedia.New York,USA:ACM,2018:274-282.
2020, Vol. 33 
