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

Quick clustering preprocessing method for massive image eigenvectors

An image feature and preprocessing technology, applied in the field of image processing, can solve the problems of complex computing and high computing memory requirements, and achieve the effect of low computing complexity and improved robustness.

Active Publication Date: 2017-03-29
SHENZHEN INTELLIFUSION TECHNOLOGIES CO LTD
View PDF5 Cites 3 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0017] In order to solve the problems in the prior art, the present invention provides a fast clustering preprocessing method in massive image feature vectors, which solves the problems of high computational memory requirements and complex calculations in the prior art

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
  • Quick clustering preprocessing method for massive image eigenvectors
  • Quick clustering preprocessing method for massive image eigenvectors

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0030] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments.

[0031]A fast clustering preprocessing method in massive image feature vectors, comprising the following steps: (A) two-level thread pool processing; (B) two-level Map storage structure processing; the two-level thread pool includes simple pre-aggregation The primary preprocessing thread pool of the class and the secondary merging thread pool for secondary clustering and merging; in the storage structure of the two-level Map, the large clustering result Map is divided into sub-Maps, and the merging and comparison operations of the sub-Maps are performed in parallel .

[0032] The primary preprocessing thread pool performs segmentation and scheduling of primary clustering tasks, specifically: the primary preprocessing thread pool performs segmentation and scheduling of primary clustering tasks, specifically: (A1) According to the upper limit of m...

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 relates to the field of image processing and discloses a quick clustering preprocessing method for massive image eigenvectors. The method comprises the following steps of (A) two-stage thread pool processing; and (B) two-stage Map storage structure processing. A two-stage thread pool comprises a primary preprocessing thread pool which performs simple pre-clustering, and a secondary combination thread pool which performs secondary clustering combination; and in a two-stage Map storage structure, a large clustering result Map is divided into sub-Maps, and combination and comparison operations of the sub-Maps are carried out in parallel. The method has the beneficial effects that a clustering operation process is suitable for running in a multi-core server through the two-stage scheduling design; and through the two-stage Map storage structure, the influence of blockage between clustering comparison and clustering update tends to 0.

Description

technical field [0001] The invention relates to the field of image processing, in particular to a fast clustering preprocessing method in massive image feature vectors. . Background technique [0002] In the field of intelligent image analysis, the early steps of image analysis generally include structured processing of images. After the structured processing process operates on the pixel matrix data of the image (such as using a neural network model), a limited-dimensional image is finally obtained. A high-dimensional vector is used to express the characteristics of the original image. This vector is generally also called the feature vector of the image, and the image feature vector is the type of input data. [0003] In the intelligent analysis of images, there are two basic problems. One is to solve the classification problem of "what" the image is (such as face detection), and the other is to solve the "who is the image individual in the known set of certain types of th...

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): G06F17/30G06K9/62
CPCG06F16/51G06F16/56G06F18/2323
Inventor 王健钟斌
Owner SHENZHEN INTELLIFUSION TECHNOLOGIES CO LTD
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