When several foreign fighters with the same type enter detection range, the electronic warfare (EW) receivers will intercept many the same type radar emitter signals. If the intercepted pulse is processed by the traditional sorting methods, the number of emitters cannot be identified. The main reason is that the same type of radar has similar parameters. It will cause a devastating influence on subsequent strategic decisions. A novel sorting method based on the trajectory features is proposed to solve the aforementioned problems. First, the trajectory features of the intercepted pulse signal are extracted. Then, the segmentation method is utilized to preprocess the signals, which enhances the computing efficiency and improves the sorting accuracy. Meanwhile, a prediction framework based on long short-term memory (LSTM) recurrent neural network is established to forecast pulses. Finally, the radar stagger pulses are sorted by forecast pulses. The simulation results show that the proposed method can recognize the number of emitters and achieve high sorting accuracy. It provides a new idea for the radar signals sorting of the same type.

HSI has abundant spectral–spatial information. Using this information to improve the accuracy of HSI classification is a hot issue in the industry. This letter proposes an end-to-end multilevel hybrid attention network (DMCN). It is composed of a dense 3-D convolutional neural network (3D-CNN), grouped residual 2D-CNN, and coordinate attention that can perceive categories. In the case of a small number of training samples, DMCN can still extract spectral–spatial fusion information and learn spatial features more deeply for classification. Experiments are conducted on three well-known hyperspectral datasets, i.e., Indian Pines (IP), University of Pavia (UP), and Salinas (SA). The results show that DMCN achieved 92.39%, 97.28%, and 98.40% classification accuracy in IP, UP, and SA.

In this paper, we propose an orientation free method for unconstrained cursive handwritten chinese word recognition. By a novel gravity center balancing method, the orientation of a handwritten word can be detected. Through the process of stroke extraction, stroke breaking, heuristic over-segmentation and path searching by recognition and lexicon information, the handwritten word with characters even connected or partially overlapped can be recognized. Experiments were performed on 173,660 unconstrained handwritten Chinese word samples collected by Pocket PC. The promising experimental results demonstrated our method is an orientation free and stroke-order free method for unconstrained cursive handwritten Chinese word recognition.

The performance of existing maneuvering target tracking methods for highly maneuvering targets in cluttered environments is unsatisfactory. This paper prop