Robust detection of infrared small target is still a challenge due to the diversity and complexity of the background. In this letter, we propose a novel detection approach based on density peaks searching and maximum-gray region growing. The main idea is that infrared small targets can be described by three features: a relatively high density, a relatively large distance from pixels with higher density, and a relatively large density gap between targets and their neighbors. This idea helps to establish a detection procedure which can detect small targets of different sizes and remove the interference caused by clutters of various complex shapes. A quartile-based technique is introduced to obtain a more robust decision threshold for multiple scenes. Compared with eight state-of-the-art algorithms, the proposed method shows a superior detection performance and an acceptable efficiency in extensive experiments.

Owing to the limitations of the imaging principle as well as the properties of imaging systems, infrared images often have some drawbacks, including low resolution, a lack of detail, and indistinct edges. Therefore, it is essential to improve infrared image quality. Considering the information of neighbors, a description of sparse edges, and by avoiding staircase artifacts, a new super-resolution reconstruction (SRR) method is proposed for infrared images, which is based on fractional order total variation (FTV) with quaternion total variation and the $L_p$ quasinorm. Our proposed method improves the sparsity exploitation of FTV, and efficiently preserves image structures. Furthermore, we adopt the plug-and-play alternating direction method of multipliers (ADMM) and the fast Fourier transform (FFT) theory for the proposed method to improve the efficiency and robustness of our algorithm; in addition, an accelerated step is adopted. Our experimental results show that the proposed method leads to excellent performances in terms of an objective evaluation and the subjective visual effect.

Upon the spreading of intelligent transportation technology, reconfigurable intelligent surface (RIS) can be applied in traditional vehicular networks to assist autonomous driving, so as to deal with blind spots and high energy consumption of long-distance communication in vehicles communication system. However, achieving bidirectional reliable communication for vehicular networks is challenging because of the high mobility of vehicles. The existing solution relies on constantly reconstructing end-to-end communication links for bidirectional communication in vehicular networks. Moreover, waiting timers on the data transmission path can double the transmission delay for data return, not to mention the potential communication delays and overheads triggered by the rapid movement of the vehicle. In order to avoid those tricky situations, this article proposes a novel two-way reliable forwarding strategy (TRFS) for vehicular named data networks and adopts the RIS symbiotic communication scheme to further reduce the transmission overhead, which has the ability to establish and maintain a reliable end-to-end communication link. On the one hand, the crucial insight of the proposed TRFS is to achieve low delay by eliminating wait timers on the data return path. Specifically, a topology-free policy is used to reduce transmission overhead and to accommodate sudden changes in network topology. On the other hand, the proposed RIS-assisted symbiotic communication scheme expands the coverage of vehicle communication and acts as a relay to compensate for the energy consumption from the long-distance transmission. Moreover, we also investigate the reliability of the proposed TRFS in vehicle communication under various vehicle densities, vehicle communication distances, and speed variations. The experiments show that the proposed TRFS scheme outperforms the classic listen first broadcast last (LFBL) fast-forwarding strategy and the novel reinforcement-based lightweight forwarding (R-LF) strategy in terms of transmission overhead, satisfaction rate, and average delay. The results reiterates the proposed RIS-assisted scheme’s ability to further reduce the transmission overhead of vehicles communication.

Many major diseases can cause changes in the morphology of blood vessels, and the segmentation of retinal blood vessels is of great significance for preventing these diseases. Obtaining complete, continuous, and high-resolution segmentation results is very challenging due to the diverse structures of retinal tissues, the complex spatial structures of blood vessels, and the presence of many small ships. In recent years, deep learning networks like UNet have been widely used in medical image processing. However, the continuous down-sampling operations in UNet can result in the loss of a significant amount of information. Although skip connections between the encoder and decoder can help address this issue, the encoder features still contain a large amount of irrelevant information that cannot be efficiently utilized by the decoder. To alleviate the irrelevant information, this paper proposes a feature selection module between the decoder and encoder that utilizes the self-attention mechanism of transformers to accurately and efficiently select the relevant encoder features for the decoder. Additionally, a lightweight Residual Global Context module is proposed to obtain dense global contextual information and establish dependencies between pixels, which can effectively preserve vascular details and segment small vessels accurately and continuously. Experimental results on three publicly available color fundus image datasets (DRIVE, CHASE, and STARE) demonstrate that the proposed algorithm outperforms existing methods in terms of both performance metrics and visual quality.