SAR Target Identification Method Based on Sample Weighted Class-specific and Shared Dictionary

Abstract

The invention discloses a SAR target identification method based on a sample weighted category-specific and shared dictionary, which mainly solves the problem of low SAR target identification performance in complex scenes in the prior art. The scheme is: 1. Extract local features for given training slices and test slices; 2. Use the local features of the training slices to obtain a global dictionary; 3. Use the global dictionary to perform standard sparse coding on the local features of the training slices and test slices respectively , to obtain the local feature encoding coefficients; 4. Perform feature merging and dimensionality reduction on the local feature encoding coefficients of the training slice and the test slice, respectively, to obtain the global features of the training slice and the global features of the test slice; 5. Use the support vector machine to analyze the test slice Global features are identified. The invention improves the identification performance and can be used for SAR target identification in complex scenes.

Description

technical field [0001] The invention belongs to the technical field of radar target identification, and mainly relates to a SAR target identification method, which can be used to provide important information for vehicle target identification and classification. Background technique [0002] Synthetic aperture radar SAR uses microwave remote sensing technology, is not affected by climate and day and night, has all-weather and all-weather working capabilities, and has the characteristics of multi-band, multi-polarization, variable viewing angle and penetrability. With the emergence of more and more airborne and spaceborne SARs, a large number of SAR data in different scenarios are brought. An important application of SAR data is automatic target recognition (ATR). Target identification in complex scenes has also become one of the current research directions. [0003] Feature extraction is an important process in the target identification process. In the past few decades, th...

AI Technical Summary

Problems solved by technology

In addition, Lincoln Laboratory also proposed several features to describe the changes in the spatial distribution of high-brightness pixels when different thresholds are added to the image. These features depend not only on the difference between the target and the background but also on the size of the target; The fourth feature is polarization features, such as purity percentage, pure even percentage and strong even percentage features, but these polarization features can only be extracted from fully polarimetric SAR data

[0004] The above-mentioned traditional features mainly have the following two shortcomings: First, these features only provide a rough and partial description of the target, and they cannot describe the detailed local shape and structure information of the target and clutter, which indicates that the discrimination cannot make full use of high-level information. Distinguish rich details in images

[0005] To sum up, with the continuous improvement of SAR image resolution, traditional features have great limitations for target identification in complex scenes.

Images