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

Multi-target segmentation defect detection method and device and computer storage medium

A defect detection and target segmentation technology, applied in computer parts, computing, image analysis, etc., can solve problems such as inability to accurately classify, inability to obtain accurate target feature parameters, and inability to meet customer real-time detection requirements.

Pending Publication Date: 2021-10-08
维库(厦门)信息技术有限公司
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] There are two types of detection methods in the existing technology, one is pure deep learning target detection and target segmentation, this method cannot obtain the precise characteristic parameters of the target, cannot detect large-resolution images in real time, and there is a large-resolution training model Problems that consume time and a lot of computing power, that is, problems that cannot meet the real-time detection requirements of customers; the other type is traditional machine learning detection, which cannot be accurately classified, requires an expert mode to adjust parameters, and the accuracy rate cannot reach 99% stably recognition rate above

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
  • Multi-target segmentation defect detection method and device and computer storage medium
  • Multi-target segmentation defect detection method and device and computer storage medium
  • Multi-target segmentation defect detection method and device and computer storage medium

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0042] This embodiment provides a kind of industrial product defect intelligent detection method of multi-target positioning and real-time target segmentation. acquisition), the image data is transmitted to the memory of the PC terminal, and the PC terminal processor of this embodiment processes the image data in the memory for the X86 processor; then, the X86 processor (processor includes but not limited to CPU, GPU, FPGA, ASIC Processor) to process the image data, such as figure 1 and figure 2 As shown, it specifically includes the following steps:

[0043] Step 1: Extract the grayscale channel of the image data to obtain the grayscale image data;

[0044] Step 2: Carry out edge detection of the upper, lower, left, and right regions of the grayscale image data and fit a straight line, wherein the edge detection adopts the Canny algorithm, and the fitted straight line algorithm adopts the HUBER loss algorithm;

[0045] Step 3: Extract the internal area according to the up...

Embodiment 2

[0061] The invention provides a multi-target positioning and real-time target segmentation intelligent detection device for industrial product defects, such as image 3 and Figure 4 shown, including:

[0062] Multi-target positioning module: used to extract gray-scale channels from image data to obtain gray-scale image data; perform edge detection in the upper, lower, left, and right regions of the gray-scale image data and fit a straight line, wherein the edge detection adopts the Canny algorithm, and the The above-mentioned fitting line algorithm adopts the HUBER loss algorithm; extract the internal area according to the up, down, left, and right lines to obtain the ROI rectangular frame to be detected; cut out the image data of the ROI rectangular frame to be detected and use the high-pass filtering algorithm and the low-pass filtering algorithm to obtain the high-pass A filter image and a low-pass filter image, wherein the high-pass filter algorithm adopts the Laplacian ...

Embodiment 3

[0068] The present invention provides a computer-readable storage medium, such as image 3 As shown, a computer program is stored thereon, and when the program is executed by a processor, the method described in the first embodiment is realized.

[0069] Those skilled in the art should understand that the embodiments of the present invention may be provided as methods, apparatuses, or computer program products. Accordingly, the present invention can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.

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 multi-target segmentation defect detection method and device, and the method comprises the steps: carrying out the gray channel extraction of image data, obtaining gray image data, carrying out the edge detection, and carrying out the fitting of a straight line; extracting an internal region according to the up-down and left-right straight lines to obtain a to-be-detected ROI rectangular frame; cutting out image data of the ROI rectangular frame to be detected, and processing the image data by using a high-pass filtering algorithm and a low-pass filtering algorithm to obtain a high-pass filtering graph and a low-pass filtering graph; carrying out differencing on the high-pass filtering image and the low-pass filtering image, taking an absolute value, and comparing the absolute value with a set threshold value to obtain binary image data with the same resolution as the ROI rectangular frame; performing Blob analysis on the binary image data to obtain Blob features; cutting out local image data in the outsourcing ROI rectangular frame of each Blob, and identifying the local image data by using a target segmentation model file obtained by a trainer to obtain an accurate target category and contour in each Blob; and carrying out geometric feature mathematical calculation on each contour.

Description

【Technical field】 [0001] The invention belongs to the technical field of industrial product detection, and specifically refers to a multi-object segmentation defect detection method, device and computer storage medium thereof. 【Background technique】 [0002] The machine vision system in the field of industrial product defect detection is mainly divided into two parts: the image acquisition unit composed of traditional cameras, lenses, light sources, camera fixing and motion mechanisms, and the machine vision detection system composed of PC hosts and image acquisition cards and other image processing units. . This kind of machine vision has high extensibility and plasticity, and it can develop customized hardware and customized software for different product defects. [0003] There are two types of detection methods in the existing technology, one is pure deep learning target detection and target segmentation, this method cannot obtain the precise characteristic parameters o...

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): G06T7/00G06T7/62G06K9/32G06K9/34G06K9/46G06N3/04
CPCG06T7/0004G06T7/62G06N3/045Y02P90/30
Inventor 林宇黄旭东
Owner 维库(厦门)信息技术有限公司
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