结合多通道模板和卷积神经网络的行人检测方法研究
A Study of Pedestrian Detection Based on Multi-Channel Template and Convolution Neural Network
DOI: 10.12677/CSA.2018.81013, PDF,   
作者: 盘先跃*, 刘怡杉, 谭 科：国防科技大学电子科学与工程学院，湖南 长沙
关键词: 通道模板；卷积神经网络；行人检测；Channel Template； Convolutional Neural Network； Pedestrian Detection
摘要: 本文在传统的依靠RGB通道模板和卷积神经网络的行人检测框架下，提出了结合多通道模板和卷积神经网络的行人检测方法。主要工作是：一是采用运动侦测方法初步检测感兴趣区域，提高行人检测效率和降低虚警。二是采用多通道模板和最大值融合策略，降低单纯依靠RGB通道模板可能存在的行人漏检现象。
Abstract: In this paper, based on the traditional pedestrian detection framework of RGB channel template and convolution neural network, a pedestrian detection method combining multi-channel tem-plate and convolution neural network is proposed. The main contribution is: firstly, the motion detection method is used to detect the interest area, improve pedestrian detection efficiency and reduce the false alarm. Secondly, multi-channel template and maximum fusion strategy are adopted to reduce the possible pedestrian leakage detection based on RGB channel template.
文章引用：盘先跃, 刘怡杉, 谭科. 结合多通道模板和卷积神经网络的行人检测方法研究[J]. 计算机科学与应用, 2018, 8(1): 97-106. https://doi.org/10.12677/CSA.2018.81013

参考文献

[1] 李权. 面向安全监控的异常声音识别的研究[D]: [硕士学位论文]. 长沙: 湖南师范大学, 2015.
[2] 张永良, 张智勤, 吴鸿韬, 等. 基于改进卷积神经网络的周界入侵检测方法[J]. 计算机科学, 2017, 44(3): 182- 186.
[3] Ribeiro, D., Nascimento, J.C., Bernardino, A., et al. (2016) Improving the Performance of Pedestrian Detectors Using Convolutional Learning. Pattern Recognition, 61, 611-619.
[4] Zhou, Z. and Xu, L. (2014) Pedestrian Detection Based on HOG and Weak-Label Structural SVM. Journal of Computational Information Systems, 10, 367-374.
[5] Tomè, D., Monti, F., Baroffio, L., et al. (2016) Deep Convolutional Neural Networks for Pedestrian Detection. Image Communication, 47, 482-489. [Google Scholar] [CrossRef] 
[6] Zhang, S., Bauckhage, C., Klein, D.A., et al. (2016) Moving Pedestrian Detection Based on Motion Segmentation. Multimedia Tools & Applications, 75, 6263-6282. [Google Scholar] [CrossRef]