Shen Zhen-kang

Efficient network training method for two-dimension DOA estimation

Training set is vary large in two-dimensional (2D) direction estimation which prevents neural network from being widely used in 2D DOA estimation. A dimension-degraded training (DDT) method is proposed in this paper to reduce the training set. In the DDT method, elevation and azimuth are estimated in two separate neural networks respectively. We construct two 1-dimensional training sets that are much smaller than the original 2-dimensional one while keeping the high resolution of angle. Simulations show the validity of the proposed method.

A real-time system for target detection and tracking in IR image sequence

This paper describes a high speed image processing system for automatic target detection and tracking in IR image sequence. To detect and track moving targets in infrared clutter in real time, we have developed an effective morphological filter which removes the noise and clutter. After obtaining the potential objects, the system calculates their centroid and labels them. The architecture of the system consists of two units: the image acquisition and displaying unit and the image processing unit. The system captures the consecutive images in real time, detects and identifies objects within three or four frames, then tracks them at the video rate. The experimental results showed the effectiveness of this proposed system.

Distinguishing point targets from decoys by fusing thermal and visual images

A method for point target recognition is presented which is based on fusing information from thermal and visual images. Usually targets and decoys are made of different materials, and there are inside heat sources in targets. Thus, their emission, absorption and reflection characteristics in infrared band and in visual band are different. The differences between targets and decoys in the two mentioned bands are used to distinguish targets from decoys. By using information from thermal and visual images, one can acquire information which can not be obtained by using thermal images alone or visual images alone.<<ETX>>

The imaging analysis of asynchronous bistatic SAR with parallel tracks

The imaging of bistatic SAR with asynchronous transceiver positions is studied in this paper. First of all, the geometry of the asynchronous transceiver model is constructed, and the approximation model of the echo from the bistatic SAR is obtained based on the analysis of the instantaneous transceiver distance. Asynchronous transceiver positions will lead to non-uniform SAR data. This paper first analyzes the method for transforming the asynchronous bistatic model into a monostatic variable motion model. Then with the non-uniform FFT, it resolves the non-uniform sampling problems induced by nonconstant velocities. The simulation experiments show that the bistatic SAR images have good quality, which verifies the validity of the method presented in this paper.

Optical flow field estimation in noise environment

Noise interference is a key problem in accurate optical flow estimation in engineering application. In this paper, firstly, the basic optical flow equation is corrected according to the noise constraint; then, the iterative arithmetic expressions of optical flow field are deduced; finally, the simulation experiments are done. The results show that the method is robust in the presence of noise.