Shan Zhang’s research while affiliated with Beihang University and other places

To efficiently detect infrared small target in the clutter background, a top-hat by the reconstruction operation-based algorithm is proposed in this chapter. Firstly, the possible target regions are extracted using top-hat by the reconstruction operation. Secondly, the remained clutter background regions are further suppressed through comparing the gray value of each pixel with a small threshold. Thirdly, the target regions are enhanced again through linear extension. Finally, the real target is detected by using the automatic thresholding method. Because the top-hat by reconstruction operation efficiently extracts real protrude image regions while some marked image regions are suppressed, the proposed algorithm is efficient for infrared small target detection.

Dim small target is an active and important research area in image processing and pattern recognition. Various algorithms have been proposed to detect and track dim small target. This paper reviews some algorithms for dim small target detection, including the wavelet based algorithms, inter-frame difference based algorithms and filter based algorithms. Also, the further development of the technologies has been briefly analyzed. (C) 2011 Published by Elsevier Ltd. Selection and/or peer-review under responsibility of [CEIS2011]

A new approach of multi-modal medical image registration is proposed to overcome the drawbacks of mutual information as taking no consideration of the space information, taking all intensities without distinction, and being sensitive to noise. The proposed method firstly extracts the phase congruencies of the reference and floating image, secondly, it computes quantitative-qualitative mutual information with the phase congruency mappings, finally, the geometric transform is optimized by Particle Swarm Optimization. The quantitative-qualitative mutual information used in our algorithm select the pixels whose utility are larger than the threshold of 1. In addition, Mutual information incorporating phase congruency assimilates the information of both intensity and space. Experiment results show that our approach is more robust in suppressing noise and can achieve higher accuracy.

We have derived a new registration method for aligning medical images using salient phase congruency and regional mutual information. The proposed method firstly extracts the salient phase congruencies of the reference and floating image; secondly, it computes the regional mutual information with the salient phase congruency mappings; finally, the geometric transform is optimized by Particle Swarm Optimization. The proposed algorithm has three features: (1) Information of both intensity and space has been taken into consideration, because regional mutual information incorporates neighborhood pixels and the phase congruency can describe the structure of an image; (2) It is invariant to contrast; (3) It is not difficult to be implemented. Experiment results show that our method has acquired a high level of accuracy and robustness compared with some other algorithms.