Image classification method, device and equipment and readable storage medium

A classification method and image technology, applied in the field of image processing, can solve the problems of time-consuming image classification processing, inability to adapt to real-time and large-scale image classification processing application scenarios, etc.

Pending Publication Date: 2019-04-19
GUANGDONG INSPUR BIG DATA RES CO LTD
View PDF7 Cites 6 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

As a result, image classification processing takes a long time and cannot adapt to real-time and large-scale image classification processing application scenarios

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 classification method, device and equipment and readable storage medium
  • Image classification method, device and equipment and readable storage medium
  • Image classification method, device and equipment and readable storage medium

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0054] Please refer to figure 1 , figure 1 It is a flowchart of an image classification method in an embodiment of the present invention, and the method includes the following steps:

[0055] S101. Acquire a target image to be classified, perform feature extraction on the target image, and obtain N different SIFT feature vectors.

[0056]Where N is a positive integer. The specific value of N is related to the number of specific key points of the target image and the calculation process in the feature extraction process, that is, different target images may correspond to the same number of SIFT feature vectors, or may correspond to different numbers of SIFT feature vectors .

[0057] The target image to be classified can be obtained by receiving an image sent by an external device, or by directly reading the target image stored in a readable storage medium. Then perform SIFT feature extraction on the target image, where SIFT, namely scale-invariant feature transform (Scale-...

Embodiment 2

[0087] In order to facilitate those skilled in the art to better understand the image classification method provided by the embodiment of the present invention, the image classification method provided by the embodiment of the present invention will be described in detail below by taking the application to a heterogeneous acceleration platform including Spark and FPGA as an example .

[0088] Spark is a mainstream computing platform in the field of big data processing, and Fpga is a computing platform in the field of heterogeneous computing. Image classification on Spark is mainly divided into two stages: the first stage is based on the BoVW model for image representation, firstly use the image features extracted from the training image (same as the target image above) to establish a visual dictionary, on this basis the image Histogram vector form represented as a dictionary. Stage 2 performs classification based on the image histogram vector. The image representation method i...

Embodiment 3

[0097] Corresponding to the above method embodiments, an embodiment of the present invention also provides an image classification device, and the image classification device described below and the image classification method described above can be referred to in correspondence.

[0098] see Figure 6 As shown, the device includes the following modules:

[0099] Feature extraction module 101, is used for obtaining target image to be classified, carries out feature extraction to target image, obtains N different SIFT feature vectors; Wherein N is a positive integer;

[0100] Clustering data sending module 102, for obtaining initial clustering center, and N SIFT feature vectors and initial clustering center are sent to FPGA;

[0101] The clustering result obtaining module 103 is used to receive the Euclidean distance calculation result returned by the FPGA, and utilize the Euclidean distance calculation result to cluster N SIFT feature vectors to obtain the clustering result; ...

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 classification method, and the method comprises the steps: obtaining a to-be-classified target image, carrying out the feature extraction of the target image, and obtaining N different SIFT feature vectors; Wherein N is a positive integer; Obtaining an initial clustering center, and sending the N SIFT feature vectors and the initial clustering center to the FPGA; Receiving an Euclidean distance calculation result returned by the FPGA, and clustering the N SIFT feature vectors by using the Euclidean distance calculation result to obtain a clustering result; Taking the clustering result as a visual dictionary, and calculating a histogram vector of the visual dictionary; And inputting the histogram vectors into a classifier for classification to obtain a classification result of the target image. According to the method, a plurality of Euclidean distances can be calculated at a time through the FPGA, so that the clustering processing speed can be shortened, and the image classification speed is further increased. The invention further discloses an image classification device and equipment and a readable storage medium which have corresponding technicaleffects.

Description

technical field [0001] The present invention relates to the technical field of image processing, in particular to an image classification method, device, equipment and readable storage medium. Background technique [0002] With the continuous development of computer technology and image acquisition methods, large-scale digital image data has been generated. Image classification technology uses computers to automatically analyze and classify images, which is one of the current research hotspots in the computer field. [0003] Among them, the visual dictionary (BoVW, Bag of Visual Words) is a commonly used image representation method, which is widely used in image classification algorithms. Because the traditional file system and processing architecture that BoVW relies on in the image classification algorithm cannot adapt to large-scale digital images, for example, in the process of clustering feature vectors to obtain a visual dictionary, clustering operations cannot be per...

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): G06K9/62G06K9/46
CPCG06V10/50G06V10/464G06F18/23213G06F18/24
Inventor 高开
Owner GUANGDONG INSPUR BIG DATA RES CO LTD
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

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap