Method for identifying handwritten numbers based on convolutional neural network and support vector machine

A convolutional neural network and support vector machine technology, applied in neural learning methods, biological neural network models, character recognition and other directions, can solve problems such as difficulty in accurately distinguishing numbers, inability to meet stability, high recognition accuracy, etc. Large error correction ability, improve prediction accuracy, and increase the effect of information volume

Inactive Publication Date: 2017-04-26
XIDIAN UNIV
View PDF2 Cites 57 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0011] 1. The strokes of the numbers are simple and the differences are relatively small, making it difficult to accurately distinguish numbers such as 3 from 8 or 5 from 6
[0012] 2. Handwritten Arabic numerals are common all over the world, and there are countless users
Compared with the current technology, most technologies cannot achieve high recognition accuracy, and are unstable and poor i

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 for identifying handwritten numbers based on convolutional neural network and support vector machine
  • Method for identifying handwritten numbers based on convolutional neural network and support vector machine
  • Method for identifying handwritten numbers based on convolutional neural network and support vector machine

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0041] With the development of information technology, numbers are closely related to people's lives. Everyone is dealing with numbers in daily life. The application of digital identification technology is becoming more and more extensive. In order to meet people's daily needs and reduce the work of Burden, the present invention has carried out innovation and research, proposes a kind of handwritten numeral recognition method based on convolutional neural network and support vector machine, see figure 1 , the handwritten digit recognition process includes the following steps:

[0042] (1): Expand the training set of handwritten digital pictures. The data set selected in this example is the MINIST data set. The sample in the MNIST data set is figure 2 As shown, the dataset is a database of handwritten digits built by Corinna Cortes of Google Labs and Yann LeCun of the Courant Institute of New York University. Will figure 2 The shown handwritten digital picture is used as th...

Embodiment 2

[0055] The handwritten digit recognition method based on convolutional neural network and support vector machine is the same as embodiment 1, wherein the hyperparameters of two networks are set in step 3, and are set as follows:

[0056] The first convolutional neural network includes 2 convolutional layers, 2 pooling layers, the size of the convolutional kernel of the first convolutional layer is 6X6, a total of 32 convolutional kernels; the size of the convolutional kernel of the second convolutional layer is 5X5, a total of 64 convolution kernels; the step size of the convolution layer is 1; the kernel size of the first pooling layer is 3X3, and the number of kernels is 32; the kernel size of the second pooling layer is 3X3, The number of cores is 64; the step size of the two pooling layers is 2; the number of neurons in the fully connected layer is 200, and the activation function of each layer is the ReLu function;

[0057] The second convolutional neural network includes...

Embodiment 3

[0062] The handwritten digit recognition method based on convolutional neural network and support vector machine is the same as embodiment 1-2, wherein the alternating structure quantity of convolutional layer and pooling layer in step 3 is 1, and in convolutional neural network, convolutional layer and The number of pooling layers is 1. When the number of convolutional layers and pooling layers is 1, the training speed of the convolutional neural network will be faster, but the recognition rate will also be reduced. If the recognition rate is not high, but the training speed is high, you can use The convolutional neural network structure scheme with the number of convolutional layers and pooling layers is 1.

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 method for identifying handwritten numbers based on a convolutional neural network (CNN) and a support vector machine (SVM). The combination of a convolutional neural network and a support vector machine increases the accuracy of identifying handwritten numbers. The method includes the following steps: enlarging a handwritten number image training set; conducting normalization; establishing two CNNS; training the two CNNs; establishing the SVM; remaining the alternating part of convolutional layers and pooling layers at the front edge of fully connected layers of the two CNNs, putting the fully connected layers of the two CNNs in serial connection and connecting the fully connected layers to the SVM to obtain a combined network; training the combined network; testing the handwritten number image testing set to obtain the result of the identification of the handwritten number. The method yields accuracy as high as 99.60%. According to the invention, the method obviates the need for complex pre-processing and is more adaptive and stable. The method also has high identification accuracy and is more reliable and robust. The method is applied to handwritten number identification in the fields of finance, postal service, data statistics, etc.

Description

technical field [0001] The invention belongs to the technical field of image processing and pattern recognition, and particularly relates to the recognition of handwritten digits, specifically a handwritten digit recognition method based on convolutional neural network and support vector machine, which is used in postal, taxation, transportation, finance and other industries. It has a very wide range of applications in practical activities. Background technique [0002] Handwritten digit recognition belongs to a category of handwritten character recognition, and handwritten recognition is divided into online handwritten recognition and offline handwritten recognition. Online handwriting recognition processes the information by recording information such as the raising and lowering of the text image, the spatial position of each pixel on the handwriting, and the time relationship between each stroke segment. During the processing, the system extracts information according to ...

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): G06K9/68G06K9/62G06N3/04G06N3/08
CPCG06N3/084G06V30/2528G06V30/248G06V30/10G06N3/045G06F18/2411
Inventor 李阳阳周林浩焦李成刘芳尚荣华马文萍马晶晶缑水平
Owner XIDIAN UNIV
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