Unlock instant, AI-driven research and patent intelligence for your innovation.

Method and device for obtaining image description information and classifier training method

An image description and description information technology, applied in the field of image processing, can solve the problem of image recognition accuracy decline, achieve the effect of accurate image description information and reduce the influence of background noise

Active Publication Date: 2015-07-29
FUJITSU LTD
View PDF6 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] However, an image often contains complex content, such as foreground objects and background parts, and it is often only for foreground objects in image recognition. There is no difference, which will lead to a decrease in the accuracy of image recognition

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 and device for obtaining image description information and classifier training method
  • Method and device for obtaining image description information and classifier training method
  • Method and device for obtaining image description information and classifier training method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0041] Embodiments of the present invention will be described below with reference to the drawings.

[0042] see figure 1 A method for obtaining image description information based on multiple images provided by an embodiment of the present invention includes:

[0043] S101: Divide multiple images to obtain sub-images of the multiple images.

[0044] In this embodiment of the present invention, the multiple images may be a training image set. When obtaining the description information of each image, the multiple images are first segmented to obtain the sub-images of the multiple images. The segmentation process can use a segmentation method to segment each image once, or use a variety of different segmentations. Multiple segmentations are performed on each image. For specific segmentation, the method of grid division can be used for segmentation, and the image can be divided into N×M rectangular areas, or triangular or polygonal grids, and the grids are not required to be u...

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

An embodiment of the invention discloses a method and a device for obtaining image description information and a classifier training method. The method for obtaining the image description information includes dividing a plurality of images, obtaining sub-images of the plurality of the images; obtaining similarity factors of the images on the basis of characteristics similarity among the sub-images; and calculating description information of the images corresponding to the sub-images on the basis of the similarity factors of the sub-images. When the description information of the images is obtained, calculation can be performed according to characteristics of the images and corresponding similarity factors of the images, characteristics of areas with large importance in the images have a large proportion, and characteristics of areas with small importance in the images have a small proportion. For example, foreground objects in the images have a large proportion in the final characteristics, and background contents have a small proportion. Influence of background noise is reduced effectively, and more accurate image description information is obtained.

Description

technical field [0001] The present invention generally relates to the technical field of image processing, in particular to a method and device for acquiring image description information, a classifier training method, and an image recognition method and device. Background technique [0002] With the rapid development of computer technology and the Internet, the number and complexity of digital images are showing an explosive growth trend in many fields. Therefore, how to quickly and effectively manage these massive images has naturally become a challenge for us, mainly including the access, storage, organization, and retrieval of these images. In order to meet this demand, since the 1990s, researchers and scholars in various fields have invested a lot of energy in the research of image recognition technology and developed some effective technologies and systems. [0003] In the image recognition system, according to the features in the image, such as color, texture and sha...

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): G06K9/64G06K9/66
Inventor 刘汝杰中村秋吾上原祐介
Owner FUJITSU LTD