Laplace regularization least square synthetic aperture radar automatic target recognition method

Abstract

The invention discloses an automatic target recognition method of Laplace regularization least square synthetic aperture radar, which relates to radar technology field and aims at improving the discrimination of SAR target image by adopting the method and get a better robustness to directions. The implementation steps in the invention are: firstly to make feature extraction for all the samples in MSTAR database by adopting KPCA, take all the data in training sets as samples with signs, and use the data in training sets as samples without signs to set up a weighted undigraph Gequals to (V, E), see the data point as the peak V of G, then define the similarity of the paired data points as the side of the graph, calculate the Laplacian of the graph and take as a regular term and added to the regularization least square and call as Laplace regularization least square to calculate the relative optimization problem. Then, classify the sample without signs by classifying function gain from exercises. The method can solve the identification problem based on the two-dimension SAR image.

Description

technical field [0001] The invention belongs to the technical field of radar, and relates to a specific application of pattern recognition technology, in particular to a Laplacian regularized least square synthetic aperture radar SAR automatic target recognition method. This method can be used to solve the recognition problem based on two-dimensional SAR images. Background technique [0002] The unique advantages of SAR technology in the detection of ground targets, especially stationary targets, and its good application prospects in the fields of modern battlefield perception and ground strikes have made the automatic target recognition technology ATR based on SAR images more and more popular. much attention. At present, scholars have conducted various researches on automatic target recognition of SAR images and proposed various methods, all of which are recognition methods based on target models. Object model description methods are generally divided into two categories:...

AI Technical Summary

Problems solved by technology

Although SVM is suitable for solving small-sample high-dimensional pattern classification problems, when grouping by 10° azimuth interval, the number of samples is too small, and each type of target has only 6-7 training samples in each azimuth unit, and the training performance is poor. In this case, a higher recognition rate cannot be achieved

In addition, this method has not undergone preprocessing of feature extraction. On the one hand, the recognition rate will be reduced due to the existence of noise. On the other hand, the purpose of dimensionality reduction cannot be achieved, which will bring a burden to the calculation.

[0006] At present, some scholars have used the KPCA feature extraction method to preprocess SAR targets, and then use SVM for target recognition. In the case of less, such as grouping by 10° azimuth interval, the training performance will also be poor

Images