Hu Bingliang

Interferogram spectrum reconstruction using modern spectral estimation

In this paper, we first discussed basic spectral estimation concepts and algorithms, then pointed out that the conventional spectrum reconstruction algorithm is virtually the same with Blackman-Tukey (BT) algorithm in spectral estimation. Then we use two different method (Auto-Regressive parametric model and eigenvector spectral estimation method) to reconstruct light spectrum from interferogram, and compared the results with conventional method for spectrum reconstruction. From the comparison we concluded that the modern spectrum estimation methods can well be adopted in Fourier Transform (FT) spectrum reconstruction and would perform excellent, especially when to get higher spectrum resolution from a short series of data recorded. Later we discussed the disadvantages and restrictions of the two methods used in the simulation. From these results and discussions, we suggested to implement various new spectral estimation method in FT spectrum reconstruction to get better performance.

Implementation of a LMS filter on FPGA employing extremeDSP and smart IP-core design

Field Programmable Gate Arrays have become popular platform for digital signal processing, however, the conventional method, which is essentially writing a source code in a hardware description language to design a product on a FPGA, is time consuming and complicated. This restricts the implementation of complex signal processing algorithm on FPGA. In this paper, two fast design methods of algorithm implementation on FPGA were proposed: ExtremeDSP solution and Smart IP-Core technology. The AccelDSP Synthesis tool, as the main component of ExtremeDSP solution allows transferring a MATLAB floating point design into a hardware module that can be implemented on a Xilinx FPGA. Employing predefined cores can cut the design time and significantly reduce risk while having access to the best performing and lowest cost component available. Implementation of a LMS-adaptive filter employing these two methods has been achieved and the simulation results indicate that both of the two fast designs have high performance.

Centroid registration difference algorithm for detecting spatial moving targets

In order to meet the needs of recognizing the targets under the moving space background by micro spacecraft target detector (MTD). According to the particular characters of the spatial background, this paper proposes an algorithm which combines the image centroid registration with the single difference method, named the Centroid Registration Difference Algorithm. In this algorithm, we labeled only the available bright sign spots and has distilled its centroid in two frames, and then used the image centroid registration to compare the two images and eliminated the same light spots between them, then the remaining bright spots in the present image were our targets. The simulation results indicate that the proposed algorithm has a low computational cost and the effect of the recognition is good.

Research on Broadband Spectral Imaging Spectrometer Based on CDP

In order to satisfy the application requirements of real-time spectral imaging for moving targets, we design a static, snapshot imaging spectrometer based on a CDP(crossed dispersion prism). The spectral imaging principle is studied, and an optical system of broadband spectral imaging spectrometer is designed according to this principle. Imaging spectrometer consists of a CDP, an imaging lens and a detector, with +/-2 degrees field of view, spectral coverage from 0.6 to 5.0 mu m, The results show that this instrument has a better ability to detect spectral from 0.6 to 5.0 mu m while the average spectral resolution is 20 nm. The technology for dynamic target real-time spectral imaging provides a new technical way. It has a great potential for detecting, locating and identifying unknown energetic events in real-time.

Coded Aperture Spectral Data Restoration Based on Two-Step Iterative Shrinkage/Threshold

Coded aperture imaging spectrometer based on the concept of compressed sensing can acquire the spectral diagram of object. Coded aperture spectral data restoration reconstructs three-dimensional data cube from two-dimensional coded image. The two-step iterative shrinkage/thresholding algorithms were derived from the iterative shrinkage threshold algorithm and weighted iterative shrinkage algorithm. Recovering coded aperture spectral data using the two-step iterative shrinkage/thresholding algorithms obtained the three-dimensional spectral data cube successfully.