Surface defect identification method based on image identification

A defect recognition and image recognition technology, applied in the field of surface defect recognition based on image recognition, can solve problems such as heavy workload, prone to false detection, missed detection, low work efficiency, etc., achieve high accuracy and reduce worker workload , the effect of fast production speed

Inactive Publication Date: 2019-04-19
HARBIN INST OF TECH
View PDF3 Cites 20 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to solve the problem of large workload and low work efficiency in the detection of the surface appearance of objects on the existing production and processing line through direct observation with the naked eye, and the situation of false detection and missed detection is prone to occur, and proposes A Surface Defect Recognition Method Based on 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
  • Surface defect identification method based on image identification
  • Surface defect identification method based on image identification
  • Surface defect identification method based on image identification

Examples

Experimental program
Comparison scheme
Effect test

specific Embodiment approach 1

[0020] The method for identifying surface defects based on image recognition in this embodiment, the method is implemented through the following steps:

[0021] Step 1. Collect the surface defect image as a sample, and expand the data of the collected surface defect image for the overall image and the detail image, increase the number of samples of the surface defect image, and obtain a complete sample of the surface defect image to improve the classification to be trained The classification effect of the model will be; wherein, the surface defect refers to the variegated points, scratches, defects, dust spots, pits, dirt, and bubbles on the surface of the object, and the surface defect image is a two-dimensional image;

[0022] Step 2, denoising the surface defect image by using the median filter algorithm as the most suitable method for denoising the surface defect image;

[0023] Step 3. Scale the surface defect image after denoising in step 2 by using the nearest neighbor ...

specific Embodiment approach 2

[0027] The difference from the first embodiment is that, in the surface defect recognition method based on image recognition in this embodiment, the overall image and the detail image of the collected surface defect image described in step 1 are data expanded, specifically:

[0028] Defect recognition is based on the feature information of normal and defect samples on the surface of the object for pattern recognition. The small number of training samples means that the feature information of the relevant object surface that can be extracted is reduced. The insufficient number of training samples is the main problem to be solved in pattern recognition. Because the location of appearance defects in the production process of the object surface is random, and the defect rate is low, the number of initial defect image samples is small, so it is necessary to expand the existing defect image samples. In the present invention, the image sample expansion is carried out respectively in t...

specific Embodiment approach 3

[0033]The difference from the specific embodiment 1 or 2 is that in the surface defect recognition method based on image recognition in this embodiment, the median filtering algorithm described in step 2 is used as the most suitable surface defect image denoising method to denoise the surface defect image process, specifically:

[0034] In the actual production line, due to the influence of imaging equipment and external environmental noise, digital images will be disturbed by noise during digitization and transmission. Effective noise reduction technology is required to improve image quality. The present invention compares the denoising effects of different denoising methods on different noises and compares the local processing effects of the original image, and selects the median filter algorithm as the most suitable method for denoising the surface image of the defective object.

[0035] Median filtering is a kind of nonlinear filtering. It is a nonlinear signal processing ...

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 surface defect identification method based on image identification. Image recognition field. The detection of the surface appearance of an object on an existing production and processing assembly line is directly observed by naked eyes, so that the problems of large workload and low working efficiency exist, and the conditions of wrong detection and missed detection are easy to occur. A surface defect identification method based on image identification comprises the following steps: collecting a surface defect image as a sample, and carrying out data expansion; Denoising the surface defect image through a median filtering algorithm; Taking the point with the maximum gray value in the four points around the sampling point in the surface defect image as the gray value of the point, and carrying out scaling processing; Extracting a contour of a defect in the surface defect image and a contour between the surface defect image and a background, converting the graylevel image into a binarized surface defect image, and performing image classification on the binarized surface defect image; The surface defect of the object can be quickly identified, and the identification accuracy is high.

Description

technical field [0001] The invention relates to an intelligent image recognition method, in particular to a surface defect recognition method based on image recognition. Background technique [0002] With the advancement of science and technology, the use of smart phones is increasing, and the penetration rate is getting higher and higher. According to the "2017-2018 China Smartphone Market Research Report" released by iiMedia Consulting, by the end of 2017, China's mobile Internet users reached 768 million, and the number of smartphone users reached 668 million. Mobile phone protective film, also known as mobile phone beauty film and mobile phone film, is a kind of cold lamination film that can be used to mount mobile phone body and screen. In China, in order to protect the mobile phone screen, more than 90% of smartphone users are used to attaching mobile phone film to their mobile phone. In 2014, the sales volume of mobile phone protective film has reached 554 million pi...

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 Applications(China)
IPC IPC(8): G06T7/00G06T5/00G06T7/13G06K9/62G01N21/89G01N21/88
CPCG06T5/002G06T7/0006G06T7/13G01N21/8851G01N21/89G06T2207/20032G01N2021/8887G06F18/2411
Inventor 林琳吕彦诚郭丰王晨钟诗胜
Owner HARBIN INST OF TECH
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