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

Method for extracting image features through low-complexity scale pyramid

An image feature, low-complexity technology, applied in character and pattern recognition, instrumentation, computing, etc., can solve the problem of high complexity of scale pyramids, reduce redundant computing, ensure retrieval performance, and save computing resources.

Active Publication Date: 2018-09-14
PEKING UNIV
View PDF5 Cites 4 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0018] Aiming at the high complexity of the scale pyramid in the feature extraction algorithm, a new scale pyramid construction method is proposed

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
  • Method for extracting image features through low-complexity scale pyramid
  • Method for extracting image features through low-complexity scale pyramid
  • Method for extracting image features through low-complexity scale pyramid

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0067] This embodiment improves the scale pyramid extraction structure; scale pyramids are generated from low scales to high scales, and low-scale filtering calculations occupy the most important complexity. A new scale pyramid construction method is designed, specifically as follows figure 2 Shown:

[0068] In order to generate the high-scale image group before the low-scale image group, the fourth image of the first group is directly filtered to generate the next group of input images. The four groups of high-scale images will be directly filtered and generated, and then through the process of feature point detection, selection, sorting, etc., a list of feature points of the high-scale group will be generated. Through the locality of feature distribution, block prediction is performed on the four ungenerated images of the first group, and image blocks with low possibility of containing feature points are skipped to generate the final first group of images. Finally, the fea...

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 extracting image features through a low-complexity scale pyramid, and the method specifically comprises the steps: filtering an original input image to generate five groups of image blocks to form the scale pyramid; performing the feature point detection of the last four groups of image blocks generated by the filtering, and obtaining a high-scale image featurepoint list; performing partitioning prediction processing of the first group of image blocks generated by filtering, and then performing the feature detection; carrying out the feature selection anddescription of the feature points detected on the first group of images, and enabling the selected feature points to be merged into the high-scale image feature point list to generate a final featurepoint list of the original input image. The method can reduce the redundant calculation in the generation process of the scale pyramid well, and the number of feature points generated by using the technology is not significantly different from the number in the conventional technology, and the retrieval performance can be well guaranteed after the feature generation. The CDVS standard can better meet the real-time requirements of feature extraction in real life needs, more computing resources are saved and the computing cost is reduced.

Description

technical field [0001] The invention relates to the technical field of computer image processing, in particular to a method for extracting image features from a low-complexity scale pyramid. Background technique [0002] MPEG-CDVS (Compact Descriptor for Visual Search, compact descriptor for visual search) is a standard algorithm for moving image search proposed by the International Motion Picture Experts Group MPEG. CDVS has a large number of applications in mobile search, such as WeChat, Google Goggles, etc. At the same time, the application scope of CDVS includes location retrieval, landmark recognition, product search and so on. CDVS performs image feature extraction and compression on the mobile terminal, and then streams the compressed data to the server for image retrieval. This avoids the stringent bandwidth requirements for directly transferring pictures for retrieval, and at the same time reduces the computing load and computing delay of the server. [0003] Wit...

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): G06K9/46
CPCG06V10/50
Inventor 贾惠柱宋嘉文李源解晓东
Owner PEKING 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