HPN-SOE: Infrared Small Target Detection and Identification Algorithm Based on Heterogeneous Parallel Networks With Similarity Object Enhancement
Shanliang Liu, Renbiao Wu, Jingyi Qu, Yunlong Li
Abstract
Infrared (IR) image small target detection and identification is a massive challenge with vast applications in essential places such as military industries and airports. Homoplastically, the unclear outline of the target in IR images, the complex image background, and the tiny percentage of targets in IR images make it challenging to accurately acquire the targets’ features and other characteristics, all of which make the research of target detection and identification in IR images significantly challenging. In this article, we first contribute an open dataset of single-frame IR small target dataset dubbed CAUC-SIRST. After that, the similarity object enhancement (SOE) module is obtained based on the input feature maps by calculating the Wasserstein distance between the local target and the local background. Then a heterogeneous parallel backbone network structure is constructed, and the feature maps obtained from three different backbone network channels are fused and stitched together. As above, the general convolution channel adaptively extracts feature information; the simple attention module (SimAM) channel increases the attention of potential targets, and the SOE channel increases the weight of targets. Finally, fusing the feature maps of the three different channels makes it possible to increase the weight of potential targets while retaining the original basic information, and this method is named heterogeneous parallel networks with SOE (HPN-SOE). Embedding the algorithm into the camera can form a sensor for IR target detection and recognition. Extensive experimental results demonstrate our outstanding performance, which outperforms other existing methods by achieving a detection accuracy of 85.7% and a detection speed of 31.2 frames/s.