Font Size: a A A

Wideband Radar, High-speed Signal Acquisition And Its One-dimensional Imaging And Feature Extraction

Posted on:2010-08-18Degree:MasterType:Thesis
Country:ChinaCandidate:S S ShenFull Text:PDF
GTID:2208360278453781Subject:Signal and Information Processing
Abstract/Summary:PDF Full Text Request
Wideband radar has played a more and more important role in today's military and commercial field, amout of researchers focus on broadband radar echo signal acquisition and feature extraction of the target. Due to the large bandwidth of the radar echo, after high-speed signal acquisition, the rate must be very high. So the follow-up work pressure of signal transmission has been aggravated. Therefore, real-time, accurate data transmission become the main point for the majority of researchers. One-dimensional range profiles of the target can provide information of physical structure and shape of the target. Obtaining High-resolution one-dimensional range profiles from the collected radar echo data and achieving the high-quality feature extraction of one-dimensional range profiles play a significant role in radar target recognition research.In this paper, with the characteristics of IF radar echo, ADC081000 as the signal acquisition device is selected. Virtex-4 FX12 produced by Xilinx Corporation is used to compile logic function by VHDL language in order to realize signal retarder storage and transmit. Finally, PCB of the whole system is completed. Based on the acquisition system, one-dimensional range profiles have been achieved. Based on the concept of equivalent scattering centers of high-resolution rang profiles, statistical features of equivalent scattering centers are extracted. Since advantages and disadvantages of equivalent scattering center's feature extraction can not be evaluated strictly. Therefore, based on the analysis of pose-sensitivity feature of range profile, the Fourier transform and bispectrum plus SVD transform method are studied in order to extract range profiles' stable characteristics and similarity of those extracted characteristics is analysed. Principal component analysis-based is applied to reduce data dimension on excellent extracted characteristics matrix. Finally the stable and low-dimensional characteristics' matrix for the follow-up study in target recognition field is obtained.
Keywords/Search Tags:High-speed signal acquisition, Feature extraction, FPGA, One-dimensional range profile
PDF Full Text Request
Related items