A dataset for infrared detection and tracking of dim-small aircraft targets under ground / air background
Bingwei Hui, Zhiyong Song, Hongqi Fan, Ping Zhong, Weidong Hu, Xiaofeng Zhang, Jianguo Ling, Hongyan Su, Wei Jin, Yongjie Zhang, Yaxi Bai
Abstract
<p indent="0mm">Infrared detection and tracking of dim-small targets has been a major content of military application studies concerning, for example, long-range precision strike, aerospace defense confrontation, battlefield intelligence and reconnaissance. However, among the means of infrared target detection and recognition, there has been either a lack of authenticity in simulated infrared data or the insufficient sample volume of measured data. To tackle this predicament, we build this dataset by focusing on the detection and tracking of low altitude flying dim-small targets, through outfield recording and subsequent data processing. It can be used in algorithm testing related to the detection of one or multiple fixed-wing UAV targets. This dataset captures varied scenarios under the background of both sky and ground. It includes 22 image sequences, 30 trajectories and 16177 frames in relation to 16944 targets. Each target corresponds to a label location in the image, and each image sequence corresponds to a label file. This dataset provides statistical bases for researches concerning infrared target characteristics, dim-small object detection and precision guidance.