Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Improved SURF algorithm based on gradient amplitude pre-operation

A gradient amplitude and algorithm technology, applied in the field of improving SURF algorithm, can solve the problems of small number of feature points and uneven feature points, and achieve the effects of uniform distribution of feature pairs, fast calculation speed, and high extraction accuracy.

Active Publication Date: 2019-08-16
HARBIN ENG UNIV
View PDF3 Cites 8 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] The object of the present invention is to provide a kind of image gradient magnitude based on underwater targets that solves the difficult problem of the small number of feature points and uneven feature points of the traditional SURF algorithm, has high feature point extraction accuracy, and has better noise suppression Improved SURF Algorithm for Precomputation

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
  • Improved SURF algorithm based on gradient amplitude pre-operation
  • Improved SURF algorithm based on gradient amplitude pre-operation
  • Improved SURF algorithm based on gradient amplitude pre-operation

Examples

Experimental program
Comparison scheme
Effect test

specific Embodiment 1

[0067] Step 1: Construct Hessian matrix, local curvature calculation.

[0068] Constructing the Hessian matrix is ​​the beginning and core operation of the SURF algorithm. The purpose of constructing the Hessian matrix is ​​to generate stable edge points of the image and lay the foundation for feature extraction. The construction method is to obtain the matrix of the second-order partial derivative of the matrix. The image operation here is to expand the form, and the variable input is expanded to obtain the partial derivative of the multivariate function. The Hessian can well describe the curvature change characteristics in the image area, so that it can generate a description. In the construction process of the image pyramid, SIFT (Scale invariant feature transform) is different from SURF here. The former uses DOG to process images to obtain relevant features, and the latter uses Hessian matrix to describe the region. The core idea is to obtain two Derivative D xx ,D xy ,...

Embodiment 1

[0115] Implementation 1, combined with the attached figure 1 , The implementation process of SURF algorithm is mainly composed of the following six parts: construction of Hessian matrix and curvature calculation, construction of scale space, image feature point location, main feature direction calculation, generation of feature descriptors and feature matching.

Embodiment 2

[0116] Implementation 2, combined with attached figure 2 There are two steps:

[0117] Step 1: Construct Hessian matrix, local curvature calculation.

[0118] Constructing the Hessian matrix is ​​the beginning and core operation of the SURF algorithm. The purpose of constructing the Hessian matrix is ​​to generate stable edge points of the image and lay the foundation for feature extraction. The construction method is to obtain the matrix of the second-order partial derivative of the matrix. The image operation here is to expand the form, and expand the variable to obtain the partial derivative of the multivariate function. The Hessian can well describe the curvature change characteristics in the image area, so that it can generate a description. In the process of constructing the image pyramid, SIFT is different from SURF here. The former uses DOG to process images to obtain relevant features, while the latter uses Hessian matrix to describe the region. The core idea is to...

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 an improved SURF algorithm based on gradient amplitude pre-operation, belongs to the field of underwater target image recognition. Based on a traditional SURF algorithm, complex Gaussian calculation is replaced through an integral graph searching mode, extreme value characteristics and high-frequency noise for describing an image exist in marginalization and sharp points ofthe image, and an image signal-to-noise ratio measurement index is introduced to highlight effective characteristics or components thereof. When a Hessian matrix is constructed through the SURF algorithm, a smooth gradient amplitude calculation method is added before Hessian, and the effect of an existing SURF algorithm is effectively improved. According to the invention, the problems of small number of feature points and non-uniform feature points of a traditional SURF algorithm are solved; the method has the advantages of being high in feature point extraction precision and better in noisesuppression force, can be introduced into underwater three-dimensional reconstruction, can effectively improve the underwater target three-dimensional reconstruction precision and quality, and provides powerful support for underwater observation and operation of an underwater robot.

Description

technical field [0001] The invention belongs to the field of underwater target image recognition, in particular to an improved SURF algorithm based on gradient amplitude pre-operation. Background technique [0002] In recent years, with the advancement of science and technology, people have begun to further explore and develop marine underwater resources. The rapid development of underwater robot technology has strongly promoted the process of human exploration of underwater space, and the research of underwater vision system based on bionic information has also been further promoted. The underwater visual measurement system has the advantages of large information bearing capacity, easy design and processing, and non-direct contact measurement, and has become a key component of many underwater equipment. Among them, the development of vision-based underwater positioning and recognition technology has important research significance and value in related fields such as robot ...

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): G06K9/46G06T7/33G06T17/00
CPCG06T7/33G06T17/00G06V10/443G06V10/462
Inventor 魏延辉杨鹏飞郑志田晨光刘静马博也牛家乐贺佳林
Owner HARBIN ENG UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com