Object Detection for Low-resolution Infrared Image in Land Battlefield Based on Deep Learning

This paper proposes a network for low-resolution infrared image target detection in the land battlefield based on YOLO v3 network. By reducing the number of pooling layers, the risk of infrared small target miss detection is reduced. In addition, when the number of infrared training samples is limited, a strategy of using visible light samples to assist the infrared image training model is proposed, which effectively reduces over-fitting. Experiments show that the proposed network has better performance in detecting small targets in infrared low-resolution images, and can effectively reduce over-fitting during model training through visible-light image-assisted training strategies.

[1]  Kaiming He,et al.  Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks , 2015, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[2]  Trevor Darrell,et al.  Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation , 2013, 2014 IEEE Conference on Computer Vision and Pattern Recognition.

[3]  Ali Farhadi,et al.  YOLO9000: Better, Faster, Stronger , 2016, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[4]  Hiromitsu Hama,et al.  Bundling Multislit-HOG Features of Near Infrared Images for Pedestrian Detection , 2009, 2009 Fourth International Conference on Innovative Computing, Information and Control (ICICIC).

[5]  Matthijs C. Dorst.  Distinctive Image Features from Scale-Invariant Keypoints , 2011 .

[6]  Wei Liu,et al.  SSD: Single Shot MultiBox Detector , 2015, ECCV.

[7]  Matti Pietikäinen,et al.  Multiresolution Gray-Scale and Rotation Invariant Texture Classification with Local Binary Patterns , 2002, IEEE Trans. Pattern Anal. Mach. Intell..

[8]  Guoliang Fan,et al.  Automatic Target Recognition in Infrared Imagery Using Dense HOG Features and Relevance Grouping of Vocabulary , 2014, 2014 IEEE Conference on Computer Vision and Pattern Recognition Workshops.

[9]  Ali Farhadi,et al.  You Only Look Once: Unified, Real-Time Object Detection , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).