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

Image feature binary coding representation method based on structure optimal subspace learning

A technology of subspace learning and image features, applied in image coding, image data processing, instruments, etc.

Inactive Publication Date: 2018-09-28
NANJING UNIV
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] Purpose of the invention: In order to solve the problems in the prior art, the present invention proposes a binary encoding representation method of image features based on structure optimal subspace learning, thereby effectively solving the problem of rapid image extraction under hash-based binary encoding data. exact search question

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
  • Image feature binary coding representation method based on structure optimal subspace learning
  • Image feature binary coding representation method based on structure optimal subspace learning
  • Image feature binary coding representation method based on structure optimal subspace learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0174] This embodiment includes the following parts:

[0175] Step 1, data preprocessing.

[0176] The workflow diagram of the data preprocessing steps is as follows: figure 2 shown.

[0177] The purpose of the feature representation and learning algorithm based on binary coding is to give a training set containing N samples where x i Represents the d-dimensional feature vector corresponding to each training sample, and uses a learning algorithm to find a set of suitable hash functions Each hash function encodes a feature vector, mapping it to a one-bit binary number. Then, learn a set of hash function combinations G(x)=[h 1 (x), h 2 (x),...,h g (x)], and use it to encode each feature Get a low-dimensional binary string, where g<<d. Compared with the original data, the encoding has a lower dimension, and its binary form can effectively reduce the data storage requirements and realize efficient storage of large-scale databases.

[0178] now assume It is a set of...

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 image feature binary coding representation method based on structure optimal subspace learning. The method comprises the steps of S1, preprocessing raw data and carrying outcomputing to obtain new data; S2, learning a projection parameter of a Hash function, wherein the projection parameter w is a parameter of a Hash function form, the parameter is defined for computingof image feature binary coding, in the step, a relaxation policy is employed to convert an original problem into the problem of learning the projection parameter w and then learning an offset parameter t, and the projection parameter w is learned in the step; and S3, learning the offset parameter t of the Hash function, in the step, influence of a sign function is taken into consideration again and is taken as compensation for relaxation limitation in the step 2.

Description

technical field [0001] The invention belongs to the field of image feature encoding, and in particular relates to an image feature binary encoding representation method based on structure optimal subspace learning. Background technique [0002] Designing fast indexing algorithms for large-scale data has a wide range of application values, such as object recognition, image retrieval, image matching, etc. When building an efficient large-scale data retrieval system, there are two main bottlenecks: data storage requirements and retrieval efficiency. In the application of image retrieval, given a query image, the user needs to retrieve images similar to it from a large-scale database, and return the results according to the similarity ranking. For this application scenario, one of the most basic methods is: firstly extract features from the query image and the database image respectively. Then, the distance between the query image and each database image is calculated accordin...

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): G06T9/00
CPCG06T9/00
Inventor 杨育彬甘元柱刘路飞张开军毛晓蛟
Owner NANJING 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