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

A memristor-based image recognition system and method

A technology of image recognition and memristor, which is applied in the field of image recognition, can solve the problems of limited integration density, high power consumption, and poor scalability, and achieve the effects of saving time, reducing energy consumption, and lowering control costs

Inactive Publication Date: 2017-04-19
HUAZHONG UNIV OF SCI & TECH
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] Aiming at the above defects or improvement needs of the prior art, the present invention provides an image recognition system and method based on memristors, the purpose of which is to use memristors to simulate neural synapses, establish a highly integrated neural network module, and use It is used in image recognition, thereby solving the technical problems of poor scalability, limited integration density and high power consumption of existing technologies

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
  • A memristor-based image recognition system and method
  • A memristor-based image recognition system and method
  • A memristor-based image recognition system and method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0068] The image recognition system based on the memristor provided by the present invention includes: an image signal extraction module, two neural network modules based on the memristor and a recognition module. The input of the recognition module is connected to the output of multiple neural network modules; the output of each neural network module is connected to the input of the recognition module, and its input is connected to the output of the signal extraction module.

[0069] The image signal extraction module is used to extract the feature vector signal of the grayscale image for training and to be recognized, and input it into the neural network module, including a host computer and 12 pulse generators. The host computer is connected with 12 pulse generators for reading grayscale image signals, extracting its eigenvectors, and delivering a plurality of eigenvalues ​​of the eigenvectors to corresponding pulse generators respectively; the pulse generators , the input ...

Embodiment 2

[0080] The first neural network module is used to represent the image model of the number "3", and the training method is as follows:

[0081] (a) Select 5 grayscale pictures representing the number "3", such as Figure 8 (a) to (e);

[0082] (b) Feedforward stage:

[0083] The image signal extraction module reads the pixel matrix of the grayscale image for each grayscale image used for training; uses the edge function as an edge detection function to detect edge pixels in the pixel matrix, and calculates the proportion of edge pixels as edge pixels Mean value; the pixel feature matrix is ​​divided into 3×4 blocks, and the ratio of edge pixels in each image block is compared with the mean value of edge pixels. If it exceeds the mean value of edge pixels, "1" is used as the feature value of the image block. Otherwise, "0" is used as the eigenvalue of the image block; the eigenvalues ​​of the image block are combined into a eigenvector in the order from top to bottom between row...

Embodiment 3

[0090] The image recognition method provided by the present invention, using the image recognition system as described in Embodiment 1, specifically includes the following steps:

[0091] (1) The image signal extraction module obtains the feature vector of the grayscale image to be recognized, and transmits the feature vector to each neural network module; the specific steps for the image signal extraction module to obtain the feature vector are as in Example 2. The image signal extraction module acquires training The specific steps of using the grayscale image feature vector.

[0092] (2) After the first neural network module and the second neural network module are trained according to the method of embodiment 2, respectively represent the model of the picture representing the number "3" and the model of the picture representing the number "5"; The first neural network module and the second neural network module respectively score the feature vectors of the image to be recog...

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 memristor based image identification system and method. The system comprises an image signal extracting module, a plurality of memristor based neural network modules, and an identification module, wherein an input end of the identification module is connected with an output end of each neural network module, and an input end of each neural network module is connected with an output end of the signal extracting module. The method includes the following steps that (1) a feature vector of a gray level image to be identified is acquired and input into the neural network modules; (2) each neural network module grades and identifies the feature vector according to an image model of each neural network module; (3) a difference value between a value of each neural network module and obtained standard values during training is calculated, and the classification of an image to be identified is judged according to the difference value. The memristor based image identification system has good expansibility, high integration density, and low power consumption; the method has low time complexity and high identification accuracy.

Description

technical field [0001] The invention belongs to the field of image recognition, and more specifically relates to a memristor-based image recognition system and method. Background technique [0002] Image recognition is a technology that uses computers to process, analyze and understand images to identify targets and objects in various patterns. Image recognition technology is of great significance to automated image processing, such as face recognition and geographic target recognition. Traditional image recognition technology is based on large-scale calculation methods, and there is an irreconcilable contradiction between the amount of calculation and the accuracy of calculation, which can no longer meet the needs. [0003] Machine learning methods, such as neural network models, are widely used in image recognition technology due to their self-learning ability to discover the rules of sample data, good fault tolerance performance, and high-speed parallel processing system...

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): G06K9/62
Inventor 陈进才周功业周可陈涛张涵周西聂昌盛缪向水卢萍
Owner HUAZHONG 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

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