A Method of Mining Salient Features from Multi-Related Images to Realize Image Retrieval

A picture feature and technology in pictures, applied in special data processing applications, instruments, electrical digital data processing, etc., to achieve the effect of improving efficiency

Inactive Publication Date: 2018-04-17
XI AN JIAOTONG UNIV
View PDF3 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to solve the problem of mining important visual content from related pictures on the mobile phone to realize image retrieval, and the mining of important visual content is usually achieved through feature matching. In view of this, the present invention proposes a method based on flexible binary description The image retrieval is realized by the matching method of descriptor features, and the flexible binary descriptor retains as much information as possible of the original features to accurately distinguish different features

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 Method of Mining Salient Features from Multi-Related Images to Realize Image Retrieval
  • A Method of Mining Salient Features from Multi-Related Images to Realize Image Retrieval
  • A Method of Mining Salient Features from Multi-Related Images to Realize Image Retrieval

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0042] The method of the present invention to mine salient features from multiple related pictures based on binary descriptors to realize image retrieval is divided into five steps: multi-correlation graph mining; generation of flexible binary descriptors; feature matching based on binary descriptors; determination Salient features; image retrieval using salient features.

[0043] 1. Multi-correlation graph mining is to find pictures related to the query graph in the user's mobile phone album. We use the classic BoW model to measure the similarity between the pictures in the album of similar users and the query graph. It consists of an offline part and an online part. The offline part includes image feature extraction, clustering, and quantization in the training set; the online part includes image features, quantization, and visual similarity calculation between images. The feature extraction method of the offline part is the same as that of the online part.

[0044] First i...

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 method for mining salient features from a plurality of related pictures to realize image retrieval, which is characterized in that the generation of flexible binary descriptors and feature matching based on the binary descriptors are used to complete the method. The flexible binary descriptor transforms each dimension of the original floating-point image feature into a binary form by comparing with the corresponding reference value of each node of the binary balanced tree; and the feature matching first calculates the feature distance based on the binary descriptor. Then the similarity scores of the normalized features are selected, and the best matching feature pairs are selected in turn, and the salient features in multiple correlation maps can be found from the best matching feature pairs. The number of salient features is small, but it represents the important visual content of the picture. Combining its geometric information for retrieval can get good retrieval results.

Description

technical field [0001] The invention relates to an image retrieval technology, in particular to a content-based image retrieval method at a mobile phone terminal. Background technique [0002] In recent years, mobile phones have experienced explosive development. According to statistics, in 2014, the number of global mobile phone users has reached 4.5 billion, and the number of smart phone users has reached 1.7 billion. For most people, especially young people, mobile phones have become an integral part of life. Compared with computers, they are more inclined to use mobile phones to do many things, such as sharing photos, checking bus routes, and especially surfing the Internet on mobile phones. With the development of smart phones, the functions of built-in cameras of mobile phones are becoming more and more powerful. It can be said that mobile phones have changed the way people take pictures. According to the statistics of Nokia Corporation in 2006, 42% of people in the...

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): G06F17/30G06K9/46G06K9/62
CPCG06F16/583G06V10/462G06V10/757
Inventor 钱学明杨锡玉
Owner XI AN JIAOTONG UNIV
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