Classification Method of Polarized SAR Image Based on Denoising Auto-encoding

Abstract

The invention discloses a polarization SAR image classification method based on a denoising automatic coding DA network, which mainly solves the problems of complicated feature extraction process, poor feature generalization ability and low classification accuracy in the prior art. The implementation steps are as follows: first input an optional polarimetric SAR image to be classified, extract the original features of the polarimetric SAR image and its neighborhood features; then logarithmically process the original features and neighborhood features to make the noise Satisfy the Gaussian distribution; secondly, determine the number of layers of the denoising automatic coding DA network, the number of nodes in each layer and the data noise and train the denoising automatic coding DA network; then use the trained denoising automatic coding DA network to treat the classified polarization SAR The image is classified to obtain the classification result of the polarimetric SAR image. The invention uses the denoising automatic encoding DA network, simplifies the process of feature extraction, improves the generalization ability of features and the classification accuracy of images, and can be used for ground object recognition of polarimetric SAR images.

Description

technical field [0001] The invention belongs to the technical field of image processing, and further relates to a data feature extraction method and a deep network classification in the technical field of polarimetric SAR image classification, which can be used for ground object recognition of polarimetric SAR images. Background technique [0002] Polarization SAR is a high-resolution active microwave remote sensing imaging radar. It has the advantages of all-weather, all-time, high resolution, and visual imaging. It can be used in many fields such as military, agriculture, navigation, and geographical surveillance. . Compared with SAR, polarimetric SAR performs full polarimetric measurement, which can obtain richer information about the target. In recent years, the classification using polarimetric SAR data has received great attention in the field of international remote sensing, and has become the main research direction of image classification. [0003] With the develo...

AI Technical Summary

Problems solved by technology

There are mainly the following supervised methods: the artificial neural network ANN classification method proposed by Heermann and Khazanie et al. (1992), the classification method based on support vector machine SVM proposed by Burges and Vapnik, these supervised methods greatly improve the calculation Speed ​​and accuracy, but has a very strong dependence on the selection of features, so it has a greater impact on the accuracy of the classification results

Images