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

Image feature-based image subject identification method

A recognition method and image feature technology, which is applied in character and pattern recognition, still image data retrieval, still image data indexing, etc., can solve the problems of search time limit, special input pictures cannot get matching results, and it is not easy to implement, etc. , to achieve the effect of improving correctness, increasing use value and search stability, and improving quality

Active Publication Date: 2017-09-01
NANJING UNIV OF SCI & TECH
View PDF5 Cites 36 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

This method encodes the image feature information and makes it into a lookup table. The more feature labels, although the search accuracy can be improved, but the search time will be limited; in the similarity matching calculation, although the determination of the threshold is based on the required The robustness is determined, but it still needs multiple trials to confirm the final confirmation
This method is still at a relatively basic stage in the analysis of image semantics, causing some special input images to not get correct matching results, and most of the time it will be restricted by the network, and it will not be easy to implement in China.

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 feature-based image subject identification method
  • Image feature-based image subject identification method
  • Image feature-based image subject identification method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0020] The present invention utilizes manually marked correct image features, through feature extraction and neural network training a mature model, marks a large knowledge base, and then retrieves the features of the input pictures to output pictures with similar features.

[0021] The specific technology can be divided into six parts:

[0022] One is the feature labeling of the training set

[0023] To label the features of the selected training set, first determine the species of the item in the picture, and refine the features according to different species, such as animals, mark the specific family, body color, eyes, and nose characteristics; after detailed labeling, it will be used as a training set to be use.

[0024] The second is the preprocessing of the test set images

[0025] Process images of different specifications to achieve a unified format. First, cut them into the most suitable size of 128*128. The object deformation generated during the cutting process is...

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 feature-based image subject identification method. According to the method, primary processing of an image is performed firstly, image features are deepened through image enhancement, and a foreground is roughly separated from a background; morphological processing is mainly used for extracting the image features, and a segmentation process is used for dividing the image into elements or target objects; the search on the image feature extraction is that the extracted image elements or target objects are represented in a numerical form suitable for subsequent processing of a computer, and finally features capable of being directly used for a classifier model generated by machine learning are formed; a distributed environment improves search efficiency and parallel computing capability; and the input image is identified through the method to obtain feature data, an image with highest similarity with the input image is searched for and output, and whether the two images are matched or not is judged. The invention provides a stable and feasible image search method. The semanteme of the image is deeply analyzed and learnt, so that the time and speed of a current search algorithm are shortened and increased, the network restrictions are avoided, and the universality is very high.

Description

technical field [0001] The invention relates to a method of how to extract image hidden information and classify a large image library for efficient identification, and output the result for use, especially a method for identifying image subjects based on image features. [0002] It has made great progress in improving retrieval efficiency, so it has broad application prospects. Background technique [0003] The image subject recognition method based on image features means that only the consumer needs to provide a picture of the desired item, and the feature information of the improved item can be obtained through this method. With the rapid growth of WWW and the rapid development of multimedia technology, it has become an urgent need for people to quickly and effectively retrieve, query and browse Internet picture information. At present, some existing image search methods mostly use text keywords and link information to search and retrieve images, and do not use the visu...

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
IPC IPC(8): G06F17/30G06K9/62
CPCG06F16/51G06F16/583G06F16/951G06F18/2415G06F18/214
Inventor 魏子涵王李娜刘继振
Owner NANJING UNIV OF SCI & TECH
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