SAR image super-resolution reconstruction method

A super-resolution reconstruction and high-resolution technology, applied in the field of synthetic aperture radar image super-resolution processing

Inactive Publication Date: 2020-02-18
HENAN UNIVERSITY
View PDF2 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] In recent years, deep learning has achieved excellent results in various fields, and its excellent feature extraction ability makes it popular in the field of image processing, but how to use convolutional neural networks to reconstruct low-resolution SAR images into high-resolution SAR images issues still to be resolved

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • SAR image super-resolution reconstruction method
  • SAR image super-resolution reconstruction method
  • SAR image super-resolution reconstruction method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0073] Step S1, SAR image preprocessing;

[0074] Specifically, as Figure 4 As shown, firstly, multi-view processing is performed on L1 level single-view complex SAR data, and the azimuth and distance multi-view are both set to 10. Then, according to the geographical information, the SAR images after multi-view processing are geocoded, and finally the refined The Lee filter processes the coherent speckle noise in the SAR image. The SAR image referred to here is the original SAR image acquired from the satellite. In the embodiment of the present invention, there is no specific requirement for the resolution of the original SAR image.

[0075] Step S2, making label set, training data set and testing data set to SAR image;

[0076] The SAR images taken in the ultra-fine strip (UFS) 3-meter resolution mode of the domestic high-three-point satellite are used as the data source of the high-resolution SAR image, or the SAR images taken in the sliding spotlight (SL) 1-meter resoluti...

Embodiment 2

[0107] Take part of the image of Danjiangkou Reservoir taken by my country's Gaofen-3 satellite in FSII mode as an example, the image resolution is 10 meters, and the image is processed. For the convenience of observation, some areas are selected for zoom-in processing, and are reconstructed by super-resolution of the present invention The model reconstructs the image. After comparison, the reconstructed image and the enlarged original image have a great improvement in clarity and interpretation ability.

[0108] In summary, the SAR image super-resolution reconstruction method of the present invention, after preprocessing the SAR image, makes a label set, a training data set and a test data set for the SAR image, cuts the SAR image to make a label set, and uses bicubic spline The image is down-sampled and cropped to obtain the training data set and the test data set, a CNN-based SAR image super-resolution reconstruction model is constructed and trained, and the low-resolution SA...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention provides an SAR image super-resolution reconstruction method. The method comprises the following steps: preprocessing SAR images, making a label set through cutting, performing downsampling on an image by utilizing bicubic splines, and performing cutting to obtain a training data set and a test data set; constructing an SAR image super-resolution reconstruction model based on the CNN; training the reconstruction model by using the label set, the training data set and the test data set; inputting the low-resolution SAR image into the trained reconstruction model to obtain a high-resolution SAR image. The convolutional neural network is used for obtaining the high-resolution image from the low-resolution image, the usability is improved, the definition and the interpretation capability of the reconstructed image are improved to a great extent, and the high-resolution SAR image reconstruction method has important significance in improving the content interpretation capability of the low-resolution image.

Description

technical field [0001] The invention relates to the technical field of synthetic aperture radar image super-resolution processing, in particular to a SAR image super-resolution reconstruction method. Background technique [0002] SAR is an active microwave remote sensing technology, which uses itself to emit electromagnetic waves to the surface, and images according to the backscatter information of electromagnetic waves. The commonly seen SAR image is an amplitude image obtained after a certain transformation, and the gray value of each pixel reflects the amplitude information of the echo collected by the imaging unit. However, the disadvantage of SAR comes from the unique imaging principle. As we all know, the larger the SAR imaging width, the lower the spatial resolution. Therefore, during the earth observation process of spaceborne SAR, the spatial resolution in most working modes is low. And due to the different phases of electromagnetic wave echoes, there will be cohe...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Applications(China)
IPC IPC(8): G06T3/40
CPCG06T3/4076G06T2207/10032G06T2207/20081G06T2207/20084
Inventor 赵建辉李宁李渊郭拯危牛世林毋琳闵林
Owner HENAN UNIVERSITY
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap