Polarimetric SAR terrain classification method based on denoising convolutional neural network

Abstract

The invention discloses a polarimetric SAR terrain classification method based on a denoising convolutional neural network. The method comprises the following steps: (1) generating a feature vector ofeach pixel point; (2) generating a training sample set and a test sample set; (3) generating a mean value graph; (4) constructing a denoising convolutional neural network; (5) training a denoising convolutional neural network; and (6) inputting the test sample into the trained denoising convolutional neural network to obtain a classification result. The polarimetric SAR terrain classification method based on the denoising convolutional neural network is adopted to classify the images, the loss of terrain information in the denoising process is reduced, more polarimetric scattering informationis reserved, and finally the classification precision is improved.

Description

technical field [0001] The invention belongs to the technical field of image processing, and further relates to a ground object classification method for Polarimetric Synthetic Aperture Radar (Polarimetric Synthetic Aperture Radar) images based on a denoising convolutional neural network in the technical field of image classification. The invention can be used for feature extraction and object classification of polarimetric SAR images. Background technique [0002] Polarimetric SAR image classification is an important step in the image interpretation process, and it is also an important research direction of polarimetric SAR image processing. Due to its strong penetrating power, polarimetric SAR can obtain a wealth of scattered surface feature information, but multiplicative noise is common, and preprocessing operations such as denoising are required before the interpretation task. It is difficult for traditional methods to balance the relationship between noise removal and...

AI Technical Summary

Problems solved by technology

The disadvantage of this method is that the use of algorithms and data domain information to preprocess polarimetric SAR data affects the efficiency of object classification

The disadvantages of this method are: First, although the sliding window operation effectively reduces the influence of multiplicative noise, the selection of the size and position of the sliding window will affect the classification accuracy of ground objects.

Since this method considers the influence of multiplicative noise in polarimetric SAR data, filtering is performed before extracting features, and the generalization ability of the convolutional neural network model is better. However, this method still has the disadvantage that the filtering The intensity is difficult to control, resulting in the loss of target information features and thus reducing the classification accuracy

Although this method is different from the traditional ground object classification technology, it uses a fully convolutional neural network model for classification, and the classification accuracy is higher. However, this method still has shortcomings. First, the polarization SAR Noise reduction processing is performed on the data, which affects the classification accuracy of ground objects

Second, the process of selecting Patch blocks reduces the efficiency of object classification

Images