Face detection has been an independent technology playing an important role in more and more fields, which makes it necessary and urgent to have its architecture reconfigurable to meet different demands on detection capabilities. This paper proposed a face detection architecture, which could be adjusted by the user according to the background, the sensor resolution, the detection accuracy and speed in different situations. This user adjustable mode makes the reconfiguration simple and efficient, and is especially suitable for portable mobile terminals whose working condition often changes frequently. In addition, this architecture could work as an accelerator to constitute a larger and more powerful system integrated with other functional modules. Experimental results show that the reconfiguration of the architecture is very reasonable in face detection and synthesized report also indicates its advantage on little consumption of area and power.

Although correlation filter-based trackers have demonstrated excellent performance for visual object tracking, there remain several challenges to be addressed. In this work, we propose a novel tracker based on the correlation filter framework. Traditional trackers face difficulty in accurately adapting to changes in the scale of the target when the target moves quickly. To address this, we suggest a scale adaptive scheme based on prediction scales. We also incorporate a speed-based adaptive model update method to further improve overall tracking performance. Experiments with samples from the OTB100 and KITTI datasets demonstrate that our method outperforms existing state-of-the-art tracking algorithms in fast motion scenes.

Aircraft landing scheduling (ALS) is one of the most important challenges in air traffic management. The target of ALS is to decide a landing scheduling sequence and calculate a landing time for each aircraft in terminal areas. These landing times are within time windows, and safety separation distances between aircraft must be kept. ALS is a complex problem, especially with a large number of aircraft. In this study, we propose a novel heuristic called CGIC to solve ALS problems. The CGIC consists of four components: a chunking rule based on costs, a landing subsequence generation rule, a chunk improvement heuristic, and a connection rule. In this algorithm, we reduce the complexity of the ALS problem by breaking it down into two or more subproblems with less aircraft. First, a feasible landing sequence is generated and divided into several subsequences as chunks by a chunking rule based on aircraft cost. Second, each chunk is regenerated by a constructive heuristic, and a perturbative heuristic is applied to improve the chunks. Finally, all chunks constitute a feasible landing sequence through a connection rule, and the landing time of each aircraft is calculated on the basis of this sequence. Simulations demonstrate that (a) the chunking rule based on cost outperforms other chunking rules based on time or weight for ALS in static instances, which have a large number of aircraft; (b) the proposed CGIC can solve the ALS problem up to 500 aircraft optimally; (c) in dynamic instances, CGIC can obtain high-quality solutions, and the computation time of CGIC is low enough to enable real-time execution.

Salient ship detection plays an important role in ensuring the safety of maritime transportation and navigation. However, due to the influence of waves, special weather, and illumination on the sea, existing saliency methods are still unable to achieve effective ship detection in a complex marine environment. To solve the problem, this paper proposed a novel saliency method based on an attention nested U-Structure (AU2Net). First, to make up for the shortcomings of the U-shaped structure, the pyramid pooling module (PPM) and global guidance paths (GGPs) are designed to guide the restoration of feature information. Then, the attention modules are added to the nested U-shaped structure to further refine the target characteristics. Ultimately, multi-level features and global context features are integrated through the feature aggregation module (FAM) to improve the ability to locate targets. Experiment results demonstrate that the proposed method could have at most 36.75% improvement in F-measure (Favg) compared to the other state-of-the-art methods.

As a kind of marine vehicles, Unmanned Surface Vehicles (USV) are widely used in military and civilian fields because of their low cost, good concealment, strong mobility and high speed. High-precision detection of obstacles plays an important role in USV autonomous navigation, which ensures its subsequent path planning. In order to further improve obstacle detection performance, we propose an encoder-decoder architecture named Fusion Refinement Network (FRN). The encoder part with a deeper network structure enables it to extract more rich visual features. In particular, a dilated convolution layer is used in the encoder for obtaining a large range of obstacle features in complex marine environment. The decoder part achieves the multiple path feature fusion. Attention Refinement Modules (ARM) are added to optimize features, and a learnable fusion algorithm called Feature Fusion Module (FFM) is used to fuse visual information. Experimental validation results on three different datasets with real marine images show that FRN is superior to state-of-the-art semantic segmentation networks in performance evaluation. And the MIoU and MPA of the FRN can peak at 97.01% and 98.37% respectively. Moreover, FRN could maintain a high accuracy with only 27.67M parameters, which is much smaller than the latest obstacle detection network (WaSR) for USV.

Detecting small infrared targets is a difficult but important task in highly cluttered coastal surveillance. The paper proposed a method called low-rank and sparse decomposition based frame difference to improve the detection performance of a surveillance system. First, the frame difference is used in adjacent frames to detect the candidate object regions which we are most interested in. Then we further exclude clutters by low-rank and sparse matrix recovery. Finally, the targets are extracted from the recovered target component by a local self-adaptive threshold. The experiment results show that, the method could effectively enhance the system's signal-to-clutter ratio gain and background suppression factor, and precisely extract target in highly cluttered coastal scene.