Industrial-video small digital identification method based on template matching and SVM

A technology of template matching and digital recognition, which is applied in character recognition, character and pattern recognition, instruments, etc., can solve the problems of not being suitable for industrial production sites, weak system specialization, slow recognition speed, etc., to improve the recognition rate and system Reliability, avoid slowing down of recognition speed, achieve simple and efficient results

Active Publication Date: 2017-07-18
XIDIAN UNIV
View PDF4 Cites 8 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the complex conditions in the actual industrial production process make the digital information in the surveillance video have the following problems: few digital pixels, digital blur, glitches, and changing industrial scenes, which leads to the inability of the previous printed digital recognition methods get good applicability
The software has powerful functions and strong versatility, but the special type of the system is not strong. In the aut

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
  • Industrial-video small digital identification method based on template matching and SVM
  • Industrial-video small digital identification method based on template matching and SVM
  • Industrial-video small digital identification method based on template matching and SVM

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0028] Existing digital recognition methods are applied to industrial scenarios, but the recognition effect of digits in videos cannot meet the requirements of industrial scenarios. The main problems are low digital recognition rate, slow recognition speed, and unrobust systems.

[0029] In order to solve the above deficiencies of the prior art, the present invention proposes a kind of industrial video digital recognition monitoring method based on template matching and support vector machine through research and experiment, see figure 1 , including the following steps:

[0030] (1) Initialization of industrial automation scene image data: it is necessary to first create a structural feature array strFeature, a training file svm.xml, and a digital sample set VecImg for the sample. Read the initialization file param.xml of the industrial automation scene, and obtain the template sample of the automation scene, the threshold threshold of binarization, the output condition parame...

Embodiment 2

[0042] The industrial video small digit recognition monitoring method based on template matching and support vector machine is the same as embodiment 1-1,

[0043] Wherein the described in step (2) carries out inverse gamma correction to input video image, comprises the following steps:

[0044] (2a) Get all rects of the current image i Region Image Mat i , to Mat i The sample is inversely corrected with gamma=0.4, and the RGB value of each pixel is adjusted to restore the image to the previous state of the display device.

[0045] (2b) Repeat step (2a) until num rects i All regions are corrected.

[0046] Because the video stream involved in the industrial scene is the sequence obtained by the acquisition card and the monitoring display device, the display device has its own gamma proofreading to adapt to the human eye observation mode; in order to obtain the true value of the data, we need to perform gamma on the images of all the areas to be recognized Anti-correction ...

Embodiment 3

[0048] The industrial video small digit recognition monitoring method based on template matching and support vector machine is the same as embodiment 1-2,

[0049] Wherein the image segmentation of the digital sample in the step (3) comprises the following steps:

[0050] (3a) For the extracted Mat i Call the minGray() adaptive binarization function to convert it into a binary image binaryImg; where the minGray() function selects the minimum value of the RGB component as its gray value and converts the input image into a gray image gray, and performs grayscale processing according to the threshold threshold Perform binarization.

[0051] (3b) Create the ImgCut class of the binary image binaryImg and call the preCut() method to obtain the horizontal and vertical projections of the numbers; the preCut() method preprocesses the binaryImg before cutting, mainly including two operations of image expansion and sharpening.

[0052] (3c) Call the Cut() method to cut the number group...

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 industrial-video small digital identification method based on template matching and SVM. The method is used to solve problems that robustness can not be guaranteed in an identification method and an identification rate is low under an industrial-production multi-scenario mode. The method comprises the following steps of carrying out gamma reverse correction on collected video images so that the images are returned to an original state; carrying out RGB minimum component binary on a sample and cutting according to horizontal vertical projection; using template matching based on a structure characteristic and SVM cascade identification based on an HOG characteristic, identifying a digital sample; and according to a production site output condition, outputting an identification result according with a demand through a serial port. Through using the method to carry out small digital identification in an industrial field monitoring video, there are advantages that self-adaption performance is good; an identification rate is high; an identification speed is fast; and operation is simple and so on. Compared to an existing identification method, by using the method of the invention, under a same industrial speed, a small digital identification rate is increased by 30%; and according to different output conditions, the method can be applied to various industrial scenes.

Description

technical field [0001] The invention belongs to the field of computer application technology, and mainly relates to a method for identifying small digits in monitoring videos in the industrial field, in particular to a method for identifying small digits in industrial videos based on template matching and SVM, which is applied to real-time monitoring in the field of industrial production. Background technique [0002] Printed digital recognition technology is an important branch of the field of pattern recognition, which has been promoted in practical applications, mainly including the following fields: license plate recognition system, electric meter digital recognition system and electronic bill recognition. At the same time, with the gradual mechanization of industrial production, many factories have also begun to use printed digital identification technology to monitor the production process. For example: at the asphalt production site, the acquisition equipment obtains ...

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): G06K9/00G06K9/62G06K9/34
CPCG06V20/41G06V30/153G06V10/758G06V2201/06G06V30/10G06F18/2411
Inventor 刘凯李玲鲍迪吕灵玥
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