angchao Li

A new algorithm for passive millimeter-wave image enhancement

Passive millimeter-wave image are heavily blurred due to the low-pass filter effect of antenna and the influence of noise. Aimed the problem, an improved method using paritial differential equations applied to passive millimeter-wave image high-resolution is proposed. Based on nonlinear bidirectional Laplace filter algorithm, a new passive millimeter-wave image enhancement is proposed. For different characteristics regions of image, using different strategies for processing. Experimental results show that the new method can improve the quality of passive millimeter-wave image and reduce the noise more effectively compared with the original algorithm.

Research on millimeter-wave image denoising method based on contourlet and compressed sensing

In this paper, we analyze the general properties of passive millimeter-wave (PMMW) imaging system, then the denoising principle of contourlet and compressed sensing is briefly described. With both denoising advantages of contourlet and compressed sensing, a PMMW image denoising algorithm is designed. And, in the process of compressed sensing denoising, a new type of observation matrix, which is called amplitude-matching variable density measurement matrix, is proposed utilizing the priority information of low-pass property of system. Experiments demonstrate that this improvement can eliminate the “scratches” parts which the contourlet denosing produces. PSNR of image is enhanced and subjective visual feeling of millimeter-wave image is obviously improved.

Data fusion of radar and IFF for aircraft identification

The problem of identification of friend-or-foe aircraft in the actual application condition is addressed. A hybrid algorithm combining fuzzy neutral network with probability factor (FNNP), multi-level fuzzy comprehensive evaluation and the Dempster-Shafer (D-S) theory is proposed. This hybrid algorithm constructs a complete process from generating the fuzzy database to the final identification, realizes the identification of friend-or-foe automatically if the training samples or experts experience can be obtained, and reduces the effect of uncertainties in the process of identification. At the same time, the whole algorithm can update the identification result with the augment of observations. The performance of the proposed algorithm is assessed by simulations. Results show that the proposed algorithm can successfully deduce the aircrafts identity even if the observations have measurement errors.

Research on a coupled shock-diffusion filter for passive millimeter-wave image enhancement

In this paper, we analyze and design a new type of diffusion process that denoises and sharpens images for resolution enhancement of passive millimeter wave images, simultaneously. In this algorithm, a coupled shock-diffusion filter is designed, where, according to image features such as edges, textures, and fine parts, it performs a forward smoothing along the tangent direction to the isophote lines, a backward sharpening along the normal directions, or only an isotropic diffusion, respectively. A qualitative evaluation of this algorithm is illustrated with simulated data as well as imaging results from the passive millimeter-wave imaging system in 91.5GHz. Experimental results demonstrate that our algorithm can substantially enhance image features on denoising image.