Infrared dim and small target detection based on two-stage U-skip context aggregation network with a missed-detection-an
- PDF / 2,416,142 Bytes
- 22 Pages / 439.642 x 666.49 pts Page_size
- 8 Downloads / 184 Views
Infrared dim and small target detection based on two-stage U-skip context aggregation network with a missed-detection-and-false-alarm combination loss Huan Wang1
· Manshu Shi1 · Hong Li2
Received: 28 December 2018 / Revised: 3 March 2019 / Accepted: 11 April 2019 / © Springer Science+Business Media, LLC, part of Springer Nature 2019
Abstract Infrared small target detection (ISTD) is a critical technique in both civil and military applications such as leak and defect inspection, cell segmentation for medicine analysis, early-warning systems and so on. Over the last decade, numerous ISTD methods have been proposed, such as methods based on image denoising, visual saliency detection, low-rank matrix recovery and traditional machine learning, but training an end-to-end deep model to detect small targets has not been fully investigated. In this regard, the paper proposes a novel deep model called UCAN for ISTD which concatenates two context aggregation networks and connects them using U-skip connections. A Missed-detection-and-False-alarm Combination(MFC) loss function, which is based on the Neyman-Pearson decision theory, is proposed to train the model and can well balance the detection rate and the false alarm rate. In addition, a two-stage detection scheme which involves a cascade of two UCANs is proposed to further improve the overall detection performance of ISTD. Extensive experiments on real infrared sequences and a single-frame image set and the comparison with state-of-the-art methods demonstrate the superiority of the proposed model. Keywords Infrared target detection · Context aggregation network · Missed-detection-and-false-alarm combination loss · Two-stage detection
1 Introduction Infrared small target detection (ISTD) is a critical technique in both civil and military domains, such as leak and defect inspection [13, 19], cell counting for medicine analysis [8], early-warning systems [12, 32, 40] and so on. Taking the early-warning system for an
Huan Wang
[email protected] 1
Nanjing University of Science and Technology, Nanjing, China
2
School of Automation, Southeast University, Nanjing, China
Multimedia Tools and Applications
example, it is required to be able to detect the incoming targets as early as possible. Since the targets of interest are much far from infrared sensors and/or they themselves are quite small, they usually occupy only a small number of pixels or even one pixel in an infrared image. Plus, the infrared radiation energy of targets decays remarkably over long distances, which makes them extremely dim. Additionally, the targets are often submerged in complex background clutters; the brightness of infrared images often changes with the varyingly dominated thermal radiation sources; and infrared images also suffer from sensor noise inevitably. In one word, robust small target detection in infrared images is still challenging. Over decades, numerous ISTD methods have been proposed. These methods can roughly be divided into two categories, Tracking-Before-Detection(TBD) [
Data Loading...
