Single-Frame Infrared Small Target (SIRST) detection aims to extract small and dim objects from the background of heavy noise and clutter. Traditional detection methods rely on handcrafted feature templates and prior knowledge, which exhibit inferior robustness in complex environments. Despite achieving promising outcomes, the existing detection methods based on CNN still suffer from the issues of small targets being lost in deep layers and long-range dependency problems. In this paper, we propose a cnn-based enhanced coordinate attention network with centralized visual center (CVC-ECANet) for infrared small target detection. Specifically, features of small target can be extracted properly by densely nested structure, where the enhanced coordinate attention module (ECAM) in nodes can enhance the features and maintain them in deep layers. In addition, by incorporating the centralized visual center module (CVCM) after the deepest layer of the network, global long-range dependencies can be captured and shallow features can be centrally modulated. Extensive experiments were conducted on the SIRST dataset, and the results demonstrate that the proposed method performs well in terms of probability of detection (Pd), false-alarm rate (Fa), and intersection of union (IoU).
In recent years, automatic target recognition (ATR) based on deep learning has achieved great success in RGB field, which has huge data support. However, due to the confidentiality of military targets, weather constraints, and high shooting costs, it is difficult to obtain a large number of real IR images which leads to the performance degradation of deep learning algorithms in IR field. This paper discusses the method of using simulation IR images as training set to get rid of dependence on the real image. However, there are still great differences between the original simulated image and the real image, which leads to many defects when using the original simulated image for training. Therefore, in this paper, we use cycleGAN to convert the original simulation image into the intermediate image closer to the real image based on generative adversarial networks (GAN). Finally, the effectiveness of this method is proved by experiments.