Method and system for super-resolution reconstruction of ship target remote sensing image

Abstract

The embodiment of the invention provides a super-resolution reconstruction method and a system of ship target remote sensing image. The method comprises the following steps of: the image in the ship target remote sensing image sample library is divided into image blocks and mapped into image block matrices so as to obtain a high-low image block mapping matrix; In accordance with that highlow imageblock mapping matrix, according to the preset network parameters, the low image block mapping matrix is trained by machine learning depth convolution neural network model, and the super-reconstructed convolution neural network weight solidification model is obtained. As that embodiment of the invention train the remote sensing image sample of a specific ship target to obtain a super-reconstructed convolution neural network weight solidification model, compared with the general model, the ship target remote sensing image processing has higher reconstruction accuracy and speed, and the outputimage resolution and clarity have been significantly improved. The ship target specific processing application can be achieved.

Description

technical field [0001] Embodiments of the present invention relate to the technical fields of image processing and machine learning, and in particular to a method and system for super-resolution reconstruction of ship target remote sensing images. Background technique [0002] Image super-resolution reconstruction technology (Super-Resolution) is an important digital image processing technology, which uses one or more low-resolution images (or motion sequences) to reconstruct a high-resolution, Images with high information content. This technology breaks through the resolution limitation of the image sensor itself, and can increase the image resolution and image quality without changing or upgrading the image acquisition hardware. The processing results are beneficial to surface visual interpretation, image target algorithm recognition, surface object type analysis, quantitative inversion accuracy improvement, and image information expression ability and utilization value. ...

AI Technical Summary

Problems solved by technology

[0004] However, at present, many image super-resolution reconstruction methods based on deep learning have their own shortcomings or deficiencies, but in general, the main problem is that there is no in-depth research on the algorithm processing of specific targets, and there is a lack of method research for specific ship targets. However, general image processing algorithms are less effective when applied to ship targets.

In addition, many super-resolution reconstruction methods mainly focus on ordinary images for super-resolution reconstruction technology research, and have poor adaptability to ship targets in optical remote sensing images.

In the research of deep convolutional neural network, the setting of network parameters has a great influence on the reconstruction accuracy, and many reconstruction methods do not give the parameter configuration suitable for the ship target, which does not have specific practical application value

Images