Unlock instant, AI-driven research and patent intelligence for your innovation.

Different-source image matching method based on template matching and twin neural network optimization

A technology of template matching and neural network, which is applied in biological neural network models, neural learning methods, image enhancement, etc., can solve the problems of heterogeneous image matching Sobel operator outline rough edges, low positioning accuracy, etc., to reduce errors and improve Accuracy, the effect of overcoming matching difficulties

Active Publication Date: 2021-05-14
JILIN UNIV
View PDF7 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] In view of the difficulties in heterogeneous image matching and the problems of thick contour edges and low positioning accuracy when the Sobel operator performs feature extraction, the present invention provides a heterogeneous image matching method based on template matching and twin neural network optimization

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
  • Different-source image matching method based on template matching and twin neural network optimization
  • Different-source image matching method based on template matching and twin neural network optimization
  • Different-source image matching method based on template matching and twin neural network optimization

Examples

Experimental program
Comparison scheme
Effect test

specific Embodiment approach 1

[0085] The feasibility of the method provided by the present invention is verified below with specific tests. The method of the present invention is compared with the existing template matching method in terms of feature point extraction, correct matching rate and matching speed.

[0086] 1. Working conditions

[0087] This experiment uses an intel core i9-9900k CPU@3.60ghz*16 processor, a PC running Windows 10, two GeForce RTX 1080Ti graphics cards, and Python as the programming language.

[0088] 2. Experimental content and result analysis

[0089] Such as image 3 As shown, by comparing the original SAR images, it can be seen that the matching image results of the present invention are more similar, and then by further outputting the upper left corner pixels of the two matching results and the actual SAR image at the upper left corner pixels of the optical image for testing and The accuracy rate of the output pixel point error is less than 5, as shown in Table 1, which p...

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 different-source image matching method based on template matching and twin neural network optimization, belongs to the field of computer image processing, and solves the problems that the current different-source image matching is difficult and the contour edge is relatively coarse and the positioning precision is not high when a Sobel operator performs feature extraction. The method comprises the following steps: preprocessing an SAR image; carrying out image half-pixel processing; carrying out sketch image conversion based on a Sobel operator; matching the templates; and carrying out fine matching on the twin neural network. The method is suitable for image processing operation of matching two different-source images with different sizes in the same area scene based on template matching. In the template matching process, sketch image conversion based on the Sobel operator is performed on the different-source images, the matching effect is enhanced, and the improved twin neural network is subsequently applied to perform fine matching, so that the matching effect of the different-source images is better, and the matching precision is higher.

Description

technical field [0001] The invention belongs to the technical field of computer image processing, and in particular relates to a heterogeneous image matching method based on template matching and twin neural network optimization. Background technique [0002] Image matching is a very important work in the field of computer vision and image processing. It is mainly used to match two or more images acquired at different times, different sensors, different viewing angles and different shooting conditions. Image matching is the basis of various image processing and applications, and the matching effect directly affects the subsequent image processing and application work. [0003] Heterogeneous image matching refers to the process of searching for the same image target by analyzing the similarity and consistency of the corresponding relationship of image content, features, structure, relationship, texture, gray scale, etc. through images acquired by different source sensors. He...

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): G06K9/62G06N3/04G06N3/08G06T3/40G06T5/00G06T5/20G06T7/13
CPCG06T3/4007G06T7/13G06T5/20G06N3/08G06T2207/10044G06V10/751G06N3/045G06T5/70
Inventor 赵岩林建宇李灵珊王世刚王学军
Owner JILIN UNIV