改进的 CNN用于单帧红外图像行人检测的方法

崔少华; 李素文; 黄金乐; 单巍

红外技术, 2020, 42 (3): 238, 网络出版: 2020-04-13

改进的 CNN用于单帧红外图像行人检测的方法

A Method of Pedestrian Detection Based on Improved CNN in Single-frame Infrared Images

论文大纲

崔少华 ^*李素文黄金乐单巍

作者单位

淮北师范大学物理与电子信息学院, 安徽淮北 235000

图像处理 LeNet-7系统单帧红外图像检测率 image processing LeNet-7 system single-frame infrared image detection rate

摘要

针对全卷积神经网络对单帧红外图像行人检测计算量大、检测率较低等问题, 提出了一种改进的 LeNet-7系统对红外图像行人检测的方法。该系统包含 3个卷积层、3个池化层, 通过错误率最小的试选法确定每层参数, 以波士顿大学建立的 BU-TIV数据库训练系统。首先, 以俄亥俄州立大学建立的 OTCBVS和 Terravic Motion IR Database红外数据库作为测试图像；然后, 采用自适应阈值的垂直和水平投影法得到感兴趣区域( regions of interest, ROI)；最后, 将得到的 ROI输入训练好的系统进行测试。3个测试集检测实验表明, 本文方法具有良好的识别能力, 与不同实验方法相比, 本文方法能有效提高检测率。

Abstract

We proposed an improved method of pedestrian detection in infrared images based on the LeNet-7 system, to address the problems of large computation and low detection rates in traditional methods based on a full convolution neural network. The system consists of three convolution layers and three pooling layers. The trail selection method with the smallest error rate is used to determine the parameters of each layer, while the BU-TIV database, established by Boston University, is used to train the system. Firstly, the Object Tracking and Classification in and Beyond the Visible Spectrum(OTCBVS) and Terravic Motion IR Database, established by Ohio State University, are used to test images. Then, the region of interest (ROI) is obtained by vertical and horizontal projection with adaptive thresholds. Finally, the ROI is input into the trained system for testing. Experiments on three test sets demonstrate that the proposed method has good recognition ability. Compared with different experimental methods, the proposed method can effectively improve the detection rate.

参考文献

[1] Nanda H , Davis L. Probabilistic template based pedestrian detection in infrared videos[C]//Intelligent Vehicle Symposium, IEEE, 2002:7712599.

[2] Bertozzi M, Broggi A, Grisleri P, et al. Pedestrian detection in infrared images[C]//IEEE Intelligent Vehicles Symposium, IEEE, 2003: 7883392.

[3] GAO Y , AI X , WANG Y , et al. U-V-Disparity based Obstacle Detection with 3D Camera and steerable filter[C]//Proc Intl Intelligent Vehicle Symposium, IEEE, 2011: 12095161.

[4] Girshick R, Donahue J, Darrell T, et al. Rich feature hierarchies for accurate object detection and semantic segmentation[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2014: 580-587.

[5] 许茗, 于晓升 , 陈东岳 , 等. 复杂热红外监控场景下行人检测 [J]. 中国图象图形学报, 2018, 23(12): 1829-1837. XU M, YU X S, CHEN D Y, et al. Man detection in complex thermal infrared monitoring scenes[J]. Chinese Journal of Image and Graphics, 2018, 23(12): 1829-1837.

[6] 谭康霞, 平鹏, 秦文虎 . 基于 YOLO模型的红外图像行人检测方法[J].激光与红外, 2018, 48(11): 1436-1442. TAN K X, PING P, QIN W H. Infrared image pedestrian detection method based on YOLO model[J]. Laser & Infrared, 2018, 48(11): 1436-1442.

[7] 陈恩加, 唐向宏, 傅博文. Faster R-CNN行人检测与再识别为一体的行人检索算法 [J].计算机辅助设计与图形学学报, 2019, 31(2): 332-339. CHEN E G, TANG X H, FU B W. Pedestrian Search Method Based on Faster R-CNN with the Integration of Pedestrian Detection and Re-identification[J]. Journal of Computer Aided Design & Graphics, 2019, 31(2): 332-339.

[8] 刘智嘉, 贾鹏, 夏寅辉 , 等. 基于红外与可见光图像融合技术发展与性能评价[J].激光与红外 , 2019, 49(5): 633-640. LIU Z J, JIA P, XIA Y H, et al. Development and performance evaluation of infrared and visible image fusion technology[J]. Laser & Infrared, 2019, 49(5): 633-640.

[9] 吴志洋, 卓勇, 李军, 等. 基于卷积神经网络的单色布匹瑕疵快速检测算法[J]. 计算机辅助设计与图形学学报, 2018, 30(12): 2262-2270. WU Z Y, ZHUO Y, LI J, et al. Fast detection algorithm of monochrome fabric defects based on convolution neural network[J]. Journal of Computer Aided Design & Graphics, 2018, 30(12): 2262-2270.

[10] 欧攀, 张正, 路奎, 等. 基于卷积神经网络的遥感图像目标检测 [J].激光与光电子学进展, 2019, 56(5): 74-80. OU P, ZHANG Z, LU K, et al. Remote sensing image target detection based on convolution neural network[J]. Laser & Optoelectronics Progress, 2019, 56(5): 74-80.

[11] Y. Lecun, L. Bottou, Y. Bengi, et al. Gradient-based learning applied to document recognition[J]. Proceedings of the IEEE, 1998, 86(11): 2278-2324.

[12] ZHENG Wu, Nathan Fuller, Diane Theriault, et al. IEEE Conference on Computer Vision and Pattern Recognition[DB/OL]. (2014-6-24) [2019-12-18]. http://csr.bu.edu/BU-TIV/BUTIV.html.

[13] 吕永标, 赵建伟, 曹飞龙. 基于复合卷积神经网络的图像去噪算法 [J].模式识别与人工智能, 2017, 30(2): 97-105. LU Y B, ZHAO J W, CAO F L. Image denoising algorithm based on compound convolution neural network[J]. Pattern Recognition & Artificial Intelligence, 2017, 30(2): 97-105.

[14] Riad I. Hammoud. OTCBVS Benchmark Dataset Collection[DB/OL]. (2014-6-22)[2019-12-18]. http://vcipl-okstate.org/pbvs/bench/.

[15] Riad I. Hammoud. Terravic Motion IR Database[DB/OL]. (2014-6-22) [20192-12-18]. http://vcipl-okstate.org/pbvs/bench/Data/05/download. html.

[16] 苏育挺, 陈耀, 吕卫. 基于近红外图像的嵌入式人员在岗检测系统 [J].红外技术, 2019, 41(4): 377-382. SU Y T, CHEN Y, LU W. Embedded on-the-job detection system based on near infrared image[J]. Infrared Technology, 2019, 41(4): 377-382.

[17] XU Y L, MA B P, HUANG R, et al. Person search in a scene by jointly modeling people commonness and person uniqueness[C]//Proceedings of the 22nd ACM International Conference on Multimedia, 2014: 937-940.

崔少华, 李素文, 黄金乐, 单巍. 改进的 CNN用于单帧红外图像行人检测的方法[J]. 红外技术, 2020, 42(3): 238. CUI Shaohua, LI Suwen, HUANG Jinle, SHAN Wei. A Method of Pedestrian Detection Based on Improved CNN in Single-frame Infrared Images[J]. Infrared Technology, 2020, 42(3): 238.

改进的 CNN用于单帧红外图像行人检测的方法

关于本站 Cookie 的使用提示

全站搜索

改进的 CNN用于单帧红外图像行人检测的方法

相关论文

相关资讯

关于本站 Cookie 的使用提示

全站搜索