A radar image speckle noise suppression method based on correlation loss convolutional neural network

A convolutional neural network and radar image technology, applied in the field of radar image speckle noise suppression based on correlation loss convolutional neural network, can solve problems such as ineffective suppression of speckle noise, insufficient adaptability, and weak model expression ability. Achieve the effect of improving speckle noise suppression effect, stable loss value, and good adaptability

Active Publication Date: 2022-02-15
HUAZHONG UNIV OF SCI & TECH
View PDF2 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] Aiming at the above defects or improvement needs of the prior art, the present invention provides a radar image speckle noise suppression method based on a correlation loss convolutional neural network, thus solving the lack of self-adaptability of the prior art and the poor expressiveness of the model. Weak, unable to effectively suppress the technical problem of speckle noise

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
  • A radar image speckle noise suppression method based on correlation loss convolutional neural network
  • A radar image speckle noise suppression method based on correlation loss convolutional neural network
  • A radar image speckle noise suppression method based on correlation loss convolutional neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0056] In order to make the object, technical solution and advantages of the present invention clearer, the present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described here are only used to explain the present invention, not to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not constitute a conflict with each other.

[0057] Such as figure 1 As shown, a radar image speckle noise suppression method based on the correlation loss convolutional neural network, the specific process of the method is:

[0058] 1. Offline training phase

[0059] 1.1. Training sample generation

[0060] Generating training samples mainly follows the following three steps:

[0061] a) Prepare several remote sensing images whose siz...

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 discloses a radar image speckle noise suppression method based on a correlation loss convolutional neural network, comprising: inputting a radar image into a trained convolutional neural network to obtain a result image suppressing speckle noise; a training method of the convolutional neural network It includes: superimposing speckle noise on the remote sensing image to obtain the superimposed speckle noise image, inputting the superimposed speckle noise image into the convolutional neural network, and obtaining the training output image; according to the mean square error of the remote sensing image and the training output image, the error loss is obtained , use the cross-correlation coefficient between the remote sensing image and the training output image to obtain the correlation loss; use the error loss and the correlation loss to perform backpropagation to update the weight parameters of the convolutional neural network; and then obtain the trained convolutional neural network. The invention utilizes the error loss and the correlation loss to update the weight parameter of the convolutional neural network, thereby effectively improving the radar image speckle noise suppression effect of the network.

Description

technical field [0001] The invention belongs to the field of radar image processing, and more specifically relates to a radar image speckle noise suppression method based on a correlation loss convolutional neural network. Background technique [0002] Radar plays a very important role in many fields of military and civilian use. Moreover, with the continuous upgrading of radar imaging technology and the improvement of radar image resolution, its application range is also expanding. [0003] However, due to its inherent radar imaging mechanism, radar images usually have speckle noise. Speckle noise is usually a kind of multiplicative noise, which is caused by mutual interference of echo signals scattered by several reflection points in each resolution unit. Speckle noise has caused great troubles to subsequent processing and information extraction, greatly affecting the practical performance of radar images. [0004] At present, in the field of radar image speckle noise su...

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 Patents(China)
IPC IPC(8): G06T5/00
CPCG06T5/002G06T2207/20084G06T2207/20081G06T2207/10044
Inventor 杨卫东徐昭良沈孔怀翟展蒋哲兴颜露新
Owner HUAZHONG UNIV OF SCI & TECH
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