Search

In recent years, the application of deep learning based on deep convolutional neural networks has gained great success in face detection. However, the large visual variations of pedestrians and faces, such as large pose variations and dark lightings, resulting in lower resolution targets, impose great challenges for these tasks in real-world applications. To solve this problem, we present a conceptually simple, end-to-end, and general framework for pedestrian and face detection. Our approach efficiently detects both pedestrian and face in an image. First, an efficient improved P-Net is developed to detect a pedestrian. Then an efficient improved R-Net1 is developed to filter pedestrian targets in the second level, and improved R-Net2 carries out the preliminary detection of face targets in the remaining pedestrian targets. In order to improve the face detection rate on a small scale, improved R-Net2 introduces a multi-level feature fusion mechanism. Last, an improved O-Net is proposed to identify pedestrian and face regions. Compared to state-of-the-art face detection methods such as Multiscale Cascade CNN、 Faceness、 Two-stage CNN、 MTCNN, the proposed method achieves promising performance on WIDER FACE benchmarks, our method also reaches promising results on the Caltech benchmarks.

As an inevitable phenomenon in most optical remote-sensing images, the effect of shadows is prominent in urban scenes. Shadow detection is critical for exploiting shadows and recovering the distorted information. Unfortunately, in general, automatic shadow detection methods for urban aerial images cannot achieve satisfactory performance due to the limitation of feature patterns and the lack of consideration of non-local contextual information. To address this challenging problem, the global-spatial-context-attention (GSCA) module was developed to self-adaptively aggregate all global contextual information over the spatial dimension for each pixel in this paper. The GSCA module was embedded into a modified U-shaped encoder–decoder network that was derived from the UNet network to output the final shadow predictions. The network was trained on a newly created shadow detection dataset, and the binary cross-entropy (BCE) loss function was modified to enhance the training procedure. The performance of the proposed method was evaluated on several typical urban aerial images. Experiment results suggested that the proposed method achieved a better trade-off between automaticity and accuracy. The F1-score, overall accuracy, balanced-error-rate, and intersection-over-union metrics of the proposed method were higher than those of other state-of-the-art shadow detection methods.

Vehicle targets in unmanned aerial vehicle (UAV) images are generally small, so a significant amount of detailed information on targets may be lost after neural computing, which leads to the poor performances of the existing recognition algorithms. Based on convolutional neural networks that utilize the YOLOv3 algorithm, this article focuses on the development of a quick automatic vehicle detection method for UAV images. First, a vehicle dataset for target recognition is constructed. Then, a novel YOLOv3 vehicle detection framework is proposed according to the following characteristics: The vehicle targets in the UAV image are relatively small and dense. The average precision (AP) increased by 5.48%, from 92.01% to 97.49%, which still remains the rather high processing speed of the YOLO network. Finally, the proposed framework is tested using three datasets: COWC, VEDAI, and CAR. The experimental results demonstrate that our method had a better detection capability.

Most practical target tracking are usually maneuvering, while most target tracking algorithm are linear filter. More estimation error is introduced from linear filter. Nowadays more and more researchers pay their attention in Maneuvering Target Tracking algorithm. Particle filter has been developed for estimation of nonlinear system states. This paper presents an improved particle filter, which can apply the maneuvering target tracking problem. In practice, the particle filter would take abundant computation for estimate the maneuvering target tracking. The ChinaGrid system use the agile and distributed federations to reduce the computing time, which achieve to fast resolution for particle filter computation of target tracing application. Lastly the simulation proves it.

The stealth coat on the surface can help target escaping from being detected by enemy radar, making the stealth target an import role on the modern war. But the coat will be damaged during the transformation, training and other situations, which can reduce the hiding peculiarity of the target, using Hilbert Huang Transform we can get the identity of the damaged target echo, realizing the stealth target coating damage detection, after improving Hilbert Huang Transform, this paper obtain a new Hilbert Huang Transform method which can be applied to detect the stealth target coating damage and obtain an effective result.

At present, application of dense targets is one of the important means of penetration. In many observation conditions, the echoes from the dense targets mixed with many aliasing signals, and conventional radar signal processing algorithms do not take the aliasing signals into account. Therefore it is difficult for conventional algorithms to recognize multi-targets. In this paper, an improved ESPRIT algorithm is proposed which can recognize the multi-targets from the aliasing echoes and greatly reduce the computational complexity without changing the algorithm accuracy, especially can obtain a better estimation in the case of low SNR environment. The proposed algorithm can firstly quickly realize the estimate of scattering center parameters of target echoes, and then based on the estimation, the aliasing targets can be recognized. The Simulation also verifies the improved ESPRIT algorithm has a better identification and recognition capability of aliasing targets in low SNR condition. Moreover because of reduction of the computational complexity, the performance of proposed algorithm is faster than conventional methods, especially in the case of multiple aliasing scattering centers.

A design of data fusion software architecture for laser radar and FLIR was introduced in the paper; this architecture can achieve data fusion on decision-layer for control and guide with multi-sensors system of guided missile. This design realizes the practicability and reliability of system enough, and collects information by laser radar and FLIR, and achieve a new method of control and guide with multi-sensors, that includes target identification, navigation, orientation, and anti-jamming.

Radar target automatic recognition is a comprehension problem. We have researched the practical radar data and concluded some valuable viewpoint. With it we put forward a whole scheme to achieve radar automatic target recognition. First the OS CAFR and binary accumulation algorithm is taken to accurately detect target. Here we lower self-adapt threshold to get more targets which maybe be false target. Then we use target trace management to remove the illusion target. The novel algorithm, called as endpoint detection based on moving window, can get more valuable data to classification. For 3D target we must build more templets in classification. So we use Neural Net as classification algorithm to reduce memory space. The simulation results prove that this scheme can get better recognition results.

This paper has presented an improved global motion estimation and complement for practical object detection. This algorithm can deal with the camera shift, rotation and scaling by using the affine model. And the two dimension motion estimation results of choosing the special blocks can be used to compute the parameter of affine model. The two dimension motion estimation can uses the former frame motion information as initialization for search. It is full-concluded the temporal redundancy. Then the algorithm has detail steps to remove the unsuitable black motion estimation and compute the parameters. Because it can accurate complement the global motion this algorithm can well be used in moving object detection and segmentation. As well as it is a fast algorithm, as a result it can be used in real-time system. Though the simulation, this algorithm can get perfect result.

Global motion estimation is still dilemma for researchers hard to determine a tradeoff between accuracy and computer complexity. This paper presents a three steps algorithm to get this tradeoff. First it uses the law of inertia for camera to quicken the black matching algorithm. The last two steps are using recursive statistic error to get better affine model parameter. The simulation results that this algorithm can deal with the camera shift, rotation and scaling, as well as taking less time.

On the moving object detection there is a new method which unites the temporal and spatial segmentation. Simultaneously this paper also proposal an automatic video object segmentation using the active contour module. The temporal segmentation adopts the reformative difference frame and phase correlation method, and the spatial segmentation uses the active contour module. This algorithm proposal a new viewpoint that only the external boundaries are the stable feather which must be extracted form the difference frame. This makes the algorithm achieve a high robust. And the same time temporal segmentation gives the active contour module an initialization of the edge points. Then the algorithm will refine the edge points through the energy equation iteratively. Though the simulation, this algorithm can get perfect result.

This paper proposes a novel algorithm for automatically and quickly extracting the external edge of the moving object. The algorithm uses both spatial and temporal information, which is abundant in image sequences and has not usually been efficiently used in conventional algorithm. Especially most attention of this algorithm is paid for the moving object external-edges. This orientation is designed to deal with the overlap of moving objects in successive frames. After the temporal segmentation the external edge of moving object is detected. Differing from current algorithms the fusing way of the temporal and spatial information has been progressed. Here the algorithm only uses spatial information of the temporal external edge to refine the edge points for getting the accurate boundary. This method will greatly decrease the compute and achieve the fast process in the real-time system. Though the simulation, this algorithm can get perfect result.