Impulsive jamming is difficult to be suppressed by traditional QR decomposition method. A sliding window QR decomposition array processing method based on hyperbolic GIVENS rotation was proposed. The input data with sliding window method can adjust adaptive weights more quickly to time-varying jamming than traditional recursive QR methods. This algorithm has high numerical stability and a modular structure, making it easy for hardware implementation. The simulation results verify that this method outperforms the QR method in the case of pulse jamming.

Military target recognition task often faces the situation of lack of training samples. In some extreme conditions, each category of military target may have only one training sample. This paper proposed a method to deal with the military aircraft recognition task with only one training sample, first the HOG and Hu moment invariant features are extracted, and then sparse based classification framework is used as the recognition algorithm. Experimental results show that the proposed method can achieve better recognition accuracy rate than the existing methods when there is only one single training sample.

In this paper，we propose a new GAN-based end-to-end image fusion network (VIFGAN) for fusion of visible and infrared images. VIFGAN contains a generator and a discriminator. We added the DenseNet module to the generator，and this module can extract deeper features and details. We also propose a Two-way regulation loss function(TWR-Loss). The loss function considers both the radiation information and texture information in the image, which can make the network suitable for image fusion tasks of different spectrum combinations. The experimental results show that in the fusion task of visible light and infrared images, the proposed network has better fusion performance than the existing fusion algorithm, the visual effect of the fused image is better, and the extracted details are more abundant.

In recent years, the rapid development of blockchain technology has gradually expanded from the financial field to the social field, bringing new opportunities and challenges for community governance innovation. In this paper, we try to find a way to use blockchain technology to solve some problems of community governance. So, a "private blockchain+consortium blockchain" concept is proposed for helping improve the capability of community governance, for the purpose of fulfilling the needs of community residents during the period of COVID-19 epidemic. This paper is concluded with the advantages and limitations of the application of blockchain technology in the community governance on epidemic prevention.

Sparse representation-based classification (SRC) has been successfully used in facial expression recognition, well-known for its high accuracy and robustness to some pollutions such as corruption and occlusion. However, the environmental change and human identity of facial expression image samples still interfere with the performance of the SRC framework and typically yields low accuracy. To improve intra-class variations' robustness, we proposed using compound-variational dictionaries to solve this issue. In addition, a two-stage SRC framework and fusion strategy has also been designed to ensure the dictionaries can achieve better performance. In the process of our work, the first stage depends on the compound dictionary, including the information of apex facial expression image. The second stage relies on a dictionary called variational dictionary constructed by the difference information between expressionless face image and apex facial expression image of one person. Subsequently, the classification results of the two stages are fused by the reconstruction error-based strategy. Due to the existing state-of-the-art SRC techniques concerning the one-stage framework, the experiment results on the Cohn-Kanada (CK +) database indicate that the accuracy of the proposed approach can be improved by about 3%. We also tested the approach on the JAFFE database and achieved nearly 100% accuracy, which verified the generalization performance of our work.

The Space-Based Infrared System (SBIRS) is an important part of the anti-missile system. As the scale of SBIRS deployment expands, it will bring unprecedented pressure to the system. In this paper, we discussed the mission of SBIRS and the timeliness, accuracy, scalability and knowledge discovery challenges, explore advanced technologies that can improve the performance of SBIRS. Six technologies were studied in this paper: compressive sensing, on-board processing, cluster-based HPC system, parallel file system, NoSQL database, and data mining technology. And we introduced the related applications of technologies mentioned above. The result can provide ideas for the improvement of SBIRS.

Early warning satellites is one of the important detective equipment of anti-missile early warning system. Learning its basic principle and common technology is indispensable for training qualified people to meet the needs of future anti-missile early warning operations all over the world. Based on analyzing the advantages of the application of flipping classroom teaching mode and the breakthrough of traditional teaching mode, this paper explores the specific teaching implementation method of flipping classroom to the course of `Principle and Technology of Early Warning Satellite', reflecting on the shortcomings, and proposing rationalization suggestions.

This paper presents an effective and efficient approach based on simulating the information processing procedure of the biological visual system to solve the occlusion problem in facial expression recognition. The proposed method is composed of three components. First, Histograms of Oriented Gradients (HOG) and Local Binary Patterns (LBP) are used to extract features, which imitate the responding to stimuli on visual cortex. Second, Sparse Representation based Classification (SRC) is used due to its robustness to occlusions. Finally, since the recognition results of HOG+SRC and LBP+SRC are complimentary because HOG mainly extracts shape information while LBP primarily represents texture information, a strategy of combining HOG+SRC and LBP+SRC is implemented. Experiments on the Cohn–Kanade database show that the proposed method achieves better performance than many existing methods, and it is robust to both random occlusions and the major component occlusions.

针对传统的稀疏编码模型中假定编码残差服从高斯分布,而在实际应用中,编码残差是不可能完全服从固定分布的问题,采用了一种鲁棒稀疏编码求解方法,并将其应用到表情识别中,通过在Cohn-Kana-da数据库上的测试,证明这一基于鲁棒稀疏编码的表情识别方法比传统基于稀疏表达的识别方法（SRC）识别率更高,并且对遮挡和噪声更鲁棒。

Recently the sparse representation based classification (SRC) is successfully used to automatically recognize facial expression, well-known for its ability to solve occlusion and corruption problems. The results of those methods which using different features conjunction with SRC framework show state of the art performance on clean or noised facial expression images. Therefore, the role of feature extraction for SRC framework will greatly affect the success of facial expression recognition (FER). In this paper, we select a new feature which called LBP map. This feature is generated using local binary pattern (LBP) operator. It is not only robust to gray-scale variation, but also extracts sufficient texture information for SRC to deal with FER problem. Then we proposed a new method using the LBP map conjunction with the SRC framework. Firstly, we compared our method with state of the art published work. Then experiments on the Cohn–Kanade database show that the LBP map + SRC can reach the highest accuracy with the lowest time-consuming on clean face images than those methods which use different features such as raw image, Downsample image, Eigenfaces, Laplacianfaces and Gabor conjunction with SRC. We also experiment the LBP map + SRC to recognize face image with partial occluded and corrupted, the result shows that this method is more robust to occlusion and corruption than existing methods based on SRC framework.

Recently, the recognition of occluded facial expressions attract more and more people’s attention . Sparse representation based classification (SRC) method gives good performance on face recognition (FR) and facial expression recognition (FER), well-known for its robustness to occlusion. Histograms of Oriented Gradient (HOG) descriptors are very efficient to represent the shape information of different facial expressions and robust to various illumination. Since, this paper propos es a novel method by using HOG descriptors conjunction with SRC framework for FER. Experiment results show that the proposed method gives better performance than the existing state-of-the-art methods. Furthermore, the proposed method is not only robust to assigned occlusion s , but also to random occlusion s .

We develop a new method to recognize facial expressions. Sparse representation based classification (SRC) is used as the classifier in this method, because of its robustness to occlusion. Histograms of Oriented Gradient (HOG) descriptors and Local Binary Patterns are used to extract features. Since the results of HOG+SRC and LBP+SRC are complimentary, we use a classifier combination strategy to fuse these two results. Experiments on Cohn-Kanade database show that the proposed method gives better performance than existing methods such as Eigen+SRC, LBP+SRC and so on. Furthermore, the proposed method is robust to assigned occlusion.

In order to guide the robots for harvesting citrus fruit, a method based on hue map and chromatic aberration map was developed to detect citrus fruits with shadow within tree canopy. Thirteen color images of natural citrus-grove scene were photoed, and the color properties of target objects were analyzed. A rule for segmenting citrus fruit from background was put forward that the pixel belongs to citrus fruits if the value of R subtracted by B is larger than the threshold value(T), and the citrus fruit with shadow can be detected correctly from the tree canopy by fusing the segmented results of hue map. The results show that the fruits with shadow can be identified integrally using the proposed method.