CN115222678A - Method and system for identifying plate material thickness in scrap steel image - Google Patents

Method and system for identifying plate material thickness in scrap steel image Download PDF

Info

Publication number
CN115222678A
CN115222678A CN202210766116.0A CN202210766116A CN115222678A CN 115222678 A CN115222678 A CN 115222678A CN 202210766116 A CN202210766116 A CN 202210766116A CN 115222678 A CN115222678 A CN 115222678A
Authority
CN
China
Prior art keywords
image
plate
scrap
edge
thickness
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202210766116.0A
Other languages
Chinese (zh)
Inventor
赵亚峰
余丹
叶琳
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Yonyou Network Technology Co Ltd
Original Assignee
Yonyou Network Technology Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Yonyou Network Technology Co Ltd filed Critical Yonyou Network Technology Co Ltd
Priority to CN202210766116.0A priority Critical patent/CN115222678A/en
Publication of CN115222678A publication Critical patent/CN115222678A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G06T7/0004—Industrial image inspection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/12—Edge-based segmentation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/60—Analysis of geometric attributes
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/60—Analysis of geometric attributes
    • G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/20—Image preprocessing
    • G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/70—Labelling scene content, e.g. deriving syntactic or semantic representations
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20112—Image segmentation details
    • G06T2207/20132—Image cropping

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Multimedia (AREA)
  • Geometry (AREA)
  • Quality & Reliability (AREA)
  • Computational Linguistics (AREA)
  • Image Analysis (AREA)

Abstract

The invention provides a method and a system for identifying the thickness of a plate type material in a steel scrap image, wherein the method for identifying the thickness of the plate type material in the steel scrap image comprises the following steps: identifying plate type materials in the scrap steel image; performing edge segmentation on the identified plate material to generate a plate material edge image; the thickness of the plate profile is calculated from the plate profile edge image. According to the method, the plate material edge image is firstly segmented, then the plate material thickness is calculated based on the segmented plate material edge image, the characteristics are extracted and the thickness is calculated to be divided into two steps, so that the plate material edge characteristics are clear, the identification is convenient, meanwhile, the calculation accuracy of the plate material thickness is ensured due to the clear plate material edge characteristics, and the accuracy of plate material grading is improved.

Description

Method and system for identifying plate material thickness in scrap steel image
Technical Field
The invention relates to the technical field of computer vision, in particular to a method and a system for identifying the thickness of a plate material in a scrap steel image.
Background
The iron and steel enterprise is in the position of lifting the weight in the manufacturing industry, and the steel scrap quality testing occupies important position in the purchase of the iron and steel enterprise, and traditional tool for testing the steel scrap quality is a ladder, and the personnel who tests the quality are carried with and climb up the ladder and climb down, carries out quality testing and grading to the steel scrap on the freight train, and is not only inefficient, but also has important safety risk.
Therefore, how to provide a scheme for identifying the thickness of the plate material based on computer vision becomes a problem to be solved urgently at present.
Disclosure of Invention
In order to solve the technical problem, a first aspect of the present invention provides a method for identifying a plate profile thickness in a scrap image.
The second aspect of the invention also provides a system for identifying the plate profile thickness in the scrap steel image.
The third aspect of the invention also provides a system for identifying the thickness of the plate profile in the scrap steel image.
The fourth aspect of the present invention also proposes a readable storage medium.
The fifth aspect of the invention also provides an intelligent scrap steel grading system.
In view of the above, the first aspect of the present invention provides a method for identifying a plate profile thickness in a scrap image, which is used in an intelligent scrap grading system, and includes: identifying plate type materials in the scrap steel image; performing edge segmentation on the identified plate material to generate a plate material edge image; and calculating the thickness of the plate type material according to the edge image of the plate type material.
According to the method for identifying the thickness of the plate type material in the steel scrap image, the obtained steel scrap image is subjected to material type identification, so that the material type of the steel scrap monomer in the steel scrap image is classified, the plate type material in the steel scrap image is identified, and the thickness of the plate type material is conveniently identified in the follow-up process; after identifying the plate material in the scrap steel image, segmenting the edge of the identified plate material to eliminate the rest interference information in the image, only segmenting the edge of the plate material to generate a plate material edge image only with the plate material edge, wherein the plate material edge image can be a binary image or an image display form which can obviously distinguish the plate material edge from other parts in the image in any other form; and calculating the thickness of the plate type material according to the obtained plate type material edge image so as to judge the plate type material according to the thickness of the plate type material. According to the method, the plate material edge image is firstly segmented, then the plate material thickness is calculated based on the segmented plate material edge image, the characteristics are extracted and the thickness is calculated to be divided into two steps, so that the plate material edge characteristics are clear, the identification is convenient, meanwhile, the calculation accuracy of the plate material thickness is ensured due to the clear plate material edge characteristics, and the accuracy of plate material grading is improved.
In addition, the method for identifying the plate type thickness in the scrap steel image in the technical scheme provided by the invention can also have the following additional technical characteristics:
in the above technical solution, before identifying the plate material in the scrap image, the method further includes: acquiring a scrap steel image; and adjusting the size of the scrap steel image to obtain the scrap steel image with the preset size.
In the technical scheme, the scrap steel image is obtained firstly, so that the subsequent identification step is conveniently carried out according to the scrap steel image, the size of the obtained scrap steel image is adjusted, the size of the scrap steel image is adjusted to be the preset size, the scrap steel image during image identification at each time is processed to be the same size, the same distribution characteristics can be obtained during subsequent image processing, meanwhile, the detection of the thickness of the plate type material according to the scrap steel image is not influenced by the size change of the image, the condition that the thickness of the plate type material in the image is changed together due to the adjustment of the size of the image is avoided, and the accuracy of the scrap steel image identification is improved.
In the above technical solution, the step of adjusting the size of the scrap image to obtain a scrap image with a preset size specifically includes: when the size of the scrap steel image is larger than the preset size, cutting the scrap steel image to the preset size; and when the size of the scrap steel image is smaller than the preset size, filling the scrap steel image to the preset size.
In the technical scheme, the size of the scrap image is adjusted to be the preset size by comparing and adjusting the size of the scrap image with the preset size, specifically, when the size of the scrap image is larger than the preset size, the scrap image is cut, the size of the scrap image is adjusted to be the preset size, when the size of the scrap image is smaller than the preset size, the scrap image is filled, the size of the scrap image is filled to be the preset size, the size of the scrap image is adjusted to be the same size, so that the same distribution of features can be obtained during subsequent image recognition, the image size is adjusted in a scaling mode, the image size adjusting mode in the application cannot change the feature size in the image, the thickness recognition of the plate material in the scrap image is more accurate and cannot be influenced by the image size adjustment, the image size adjusting mode in the scaling mode directly can cause the feature size in the image to change, the calculation precision of the thickness can be influenced, the error of the calculation result is larger, and the accuracy of the cutting of the scrap image is ensured by the method of the image size adjustment.
In the above technical solution, the step of performing edge segmentation on the identified plate material to generate a plate material edge image specifically includes: and performing edge segmentation on the plate material in the scrap steel image according to the semantic segmentation model to generate a plate material edge image.
In the technical scheme, edge segmentation is carried out on the plate material in the scrap steel image through the semantic segmentation model, so that a plate material edge image is generated, the calculation of the thickness of the plate material is carried out based on the plate material edge image, meanwhile, the semantic segmentation model can well combine the global and local characteristics in the scrap steel image, the edge of the plate material of the main body in the scrap steel image is segmented, the segmentation result is more accurate, the interference information in the image can be eliminated, and the segmentation result is more accurate.
In the above technical solution, the step of calculating the thickness of the sheet material according to the edge image of the sheet material specifically includes: calculating the perimeter of the plate molding material edge and the area of the plate molding material edge in the plate molding material edge image; and obtaining the thickness of the plate profile according to the perimeter and the area of the edge of the plate profile.
According to the technical scheme, the perimeter and the area of the edge of the plate type material in the image of the edge of the plate type material are calculated, so that the thickness of the plate type material is calculated according to the calculated perimeter and area results, and the plate type material is judged according to the calculated thickness of the plate type material. The specific thickness calculation mode can be understood as moving the pixel point position in the picture to make the shape of the pixel point position be a regular shape, and since the perimeter and the area of the whole graph are not changed in the pixel point moving process, the calculation of the thickness becomes simple after the pixel point position is adjusted to obtain a regular graph (the calculation is only convenient to understand and does not represent the actual calculation process).
In the above technical solution, the step of calculating the perimeter of the plate profile edge and the area of the plate profile edge in the plate profile edge image specifically includes: calculating the perimeter of the plate molding material edge in the plate molding material edge image through 8-field algorithm; and calculating the area of the plate profile edge in the plate profile edge image by using a threshold value method.
In the technical scheme, the perimeter and the thickness of the edge of the plate type material are respectively calculated through an 8-field algorithm and a threshold value method, so that the thickness of the plate type material is conveniently calculated according to the relationship between the perimeter and the area of the edge of the plate type material, and the plate type material is judged according to the thickness of the plate type material.
A second aspect of the present invention provides a system for identifying a plate profile thickness in a scrap image, comprising: the identification module is used for identifying the plate type materials in the scrap steel image; the segmentation module is used for performing edge segmentation on the identified plate material to generate a plate material edge image; and the calculating module is used for calculating the thickness of the plate material according to the edge image of the plate material.
The system for identifying the thickness of the plate type material in the scrap steel image provided by the technical scheme of the invention comprises a feature identification module, a segmentation module and a calculation module. The identification module is used for identifying the plate type materials in the scrap steel image; the segmentation module is used for performing edge segmentation on the identified plate material to generate a plate material edge image; the calculating module is used for calculating the thickness of the plate material according to the plate material edge image. Meanwhile, according to the system for identifying the plate type thickness in the steel scrap image provided by the technical scheme of the invention, as the system is used for realizing the steps of the method for identifying the plate type thickness in the steel scrap image provided by the first aspect of the invention, the system for identifying the plate type thickness in the steel scrap image has all the technical effects of the method for identifying the plate type thickness in the steel scrap image, and the description is omitted here.
In the above technical solution, the system for identifying the plate profile thickness in the scrap image further comprises: and the preprocessing module is used for acquiring the scrap steel image and adjusting the size of the scrap steel image to obtain the scrap steel image with the preset size.
In the technical scheme, the scrap steel image is obtained firstly, so that the subsequent identification step is conveniently carried out according to the scrap steel image, the size of the obtained scrap steel image is adjusted, the size of the scrap steel image is adjusted to be the preset size, the scrap steel image during image identification at each time is processed to be the same size, the same distribution characteristics can be obtained during subsequent image processing, meanwhile, the detection of the thickness of the plate type material according to the scrap steel image is not influenced by the size change of the image, the condition that the thickness of the plate type material in the image is changed together due to the adjustment of the size of the image is avoided, and the accuracy of the scrap steel image identification is improved.
In the above technical solution, the preprocessing module is specifically configured to: when the size of the scrap steel image is larger than the preset size, cutting the scrap steel image to the preset size; and when the size of the scrap steel image is smaller than the preset size, filling the scrap steel image to the preset size.
In the technical scheme, the size of the scrap image is adjusted to be the preset size by comparing and adjusting the size of the scrap image with the preset size, specifically, when the size of the scrap image is larger than the preset size, the scrap image is cut, the size of the scrap image is adjusted to be the preset size, when the size of the scrap image is smaller than the preset size, the scrap image is filled, the size of the scrap image is filled to be the preset size, the size of the scrap image is adjusted to be a uniform size, so that the same distribution of features can be obtained during subsequent image recognition, and the adjustment of the picture size is performed in a zooming mode.
In the foregoing technical solution, the segmentation module is specifically configured to: and performing edge segmentation on the plate material in the scrap steel image according to the semantic segmentation model to generate a plate material edge image.
In the technical scheme, edge segmentation is carried out on the plate material in the scrap steel image through the semantic segmentation model, so that a plate material edge image is generated, the calculation of the thickness of the plate material is carried out based on the plate material edge image, meanwhile, the semantic segmentation model can well combine the global and local characteristics in the scrap steel image, the edge of the plate material of the main body in the scrap steel image is segmented, the segmentation result is more accurate, the interference information in the image can be eliminated, and the segmentation result is more accurate.
In the above technical solution, the calculation module is specifically configured to: calculating the perimeter of the plate molding material edge and the area of the plate molding material edge in the plate molding material edge image; and obtaining the thickness of the plate profile according to the perimeter and the area of the edge of the plate profile.
According to the technical scheme, the perimeter and the area of the edge of the plate type material in the image of the edge of the plate type material are calculated, so that the thickness of the plate type material is calculated according to the calculated perimeter and area results, and the plate type material is judged according to the calculated thickness of the plate type material. The specific thickness calculation method can be understood as moving the pixel point position in the picture to make the shape of the pixel point position be a regular shape, and since the perimeter and the area of the whole graph are not changed in the pixel point moving process, the calculation of the thickness becomes simple after the pixel point position is adjusted to obtain a regular graph (the calculation is only convenient to understand and does not represent the actual calculation process).
In the above technical solution, the calculation module is further specifically configured to: calculating the perimeter of the plate molding material edge in the plate molding material edge image through 8-field algorithm; and calculating the area of the plate type material edge in the plate type material edge image by a threshold value method.
In the technical scheme, the perimeter and the thickness of the edge of the plate type material are respectively calculated through an 8-field algorithm and a threshold value method, so that the thickness of the plate type material is calculated according to the relationship between the perimeter and the area of the edge of the plate type material, and the plate type material is judged according to the thickness of the plate type material.
A third aspect of the present invention provides a system for identifying plate type thickness in a steel scrap image, including a memory and a processor, where the memory stores a program or instructions executable on the processor, and the program or instructions, when executed by the processor, implement the steps of the method for identifying plate type thickness in a steel scrap image according to any one of the above technical solutions.
The system for identifying the plate section thickness in the scrap image provided by the technical scheme of the invention comprises a memory, a processor and a program which is stored on the memory and can run on the processor, wherein the program realizes the steps defined by any one of the methods for identifying the plate section thickness in the scrap image when being executed by the processor. Meanwhile, the system for identifying the thickness of the plate type material in the steel scrap image can realize the step limited by any method for identifying the thickness of the plate type material in the steel scrap image, so that the system for identifying the thickness of the plate type material in the steel scrap image provided by the technical scheme has all the beneficial effects of the method for identifying the thickness of the plate type material in the steel scrap image provided by any technical scheme.
A fourth aspect of the present invention provides a readable storage medium, on which a program and/or instructions are stored, which when executed by a processor, implement the steps of the method for identifying the thickness of the slab in the scrap image according to any of the above technical solutions.
According to the readable storage medium provided by the technical solution of the present invention, since the program and/or the instructions stored thereon are executed by the processor, the steps of the method for identifying the plate type thickness in the steel scrap image in any one of the above technical solutions can be implemented, so that all the beneficial technical effects of the method for identifying the plate type thickness in the steel scrap image are provided, and details are not repeated herein.
The fifth aspect of the invention provides an intelligent scrap steel grading system, which comprises a system for identifying the thickness of a plate profile in a scrap steel image according to any one of the above technical schemes; or a readable storage medium as in the above-mentioned solutions.
According to the intelligent steel scrap grading system provided by the technical scheme of the invention, the intelligent steel scrap grading system comprises the system for identifying the thickness of the plate type material in the steel scrap image in any technical scheme or the readable storage medium in the technical scheme. Therefore, the intelligent scrap steel grading system has all the technical effects of the system for identifying the thickness of the plate profile in the scrap steel image or the readable storage medium, and the details are not repeated herein.
Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention.
Drawings
The above and/or additional aspects and advantages of the present invention will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings of which:
FIG. 1 is a schematic flow chart of a method of identifying plate profile thickness in a scrap image according to an embodiment of the present invention;
FIG. 2 is a block diagram of a system for identifying plate profile thickness in a scrap image according to an embodiment of the present invention;
FIG. 3 is a block diagram of a system for identifying slab thickness in scrap images according to an embodiment of the present invention;
FIG. 4 is a flow chart illustrating a method of identifying plate profile thickness in a scrap image according to another embodiment of the present invention;
FIG. 5 is a schematic representation of a scrap image according to another embodiment of the present invention;
FIG. 6 is a schematic illustration of a pre-processed scrap image according to another embodiment of the present invention;
FIG. 7 is a schematic diagram of a binary image of a sheet stock edge according to another embodiment of the invention.
Wherein, the correspondence between the reference numbers and the component names in fig. 2 and 3 is:
200 a system for identifying plate section thickness in a scrap image, 202 an identification module, 204 a segmentation module, 206 a calculation module, 300 a system for identifying plate section thickness in a scrap image, 302 a memory, 304 a processor.
Detailed Description
In order that the above objects, features and advantages of the present invention can be more clearly understood, a more particular description of the invention will be rendered by reference to the appended drawings. It should be noted that the embodiments and features of the embodiments of the present application may be combined with each other without conflict.
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention, however, the present invention may be practiced otherwise than as specifically described herein and, therefore, the scope of the present invention is not limited by the specific embodiments disclosed below.
Methods and systems for identifying plate profile thickness in scrap images in some embodiments of the invention are described below with reference to fig. 1-7.
An embodiment of the first aspect of the present invention provides a method for identifying a plate profile thickness in a scrap image, as shown in fig. 1, including:
s102, identifying plate type materials in the scrap steel image;
s104, performing edge segmentation on the identified plate type material to generate a plate type material edge image;
and S106, calculating the thickness of the plate type material according to the edge image of the plate type material.
According to the method for identifying the plate section thickness in the scrap steel image, the obtained scrap steel image is subjected to material type identification, so that the scrap steel monomers in the scrap steel image are subjected to material type classification, and the plate section material in the scrap steel image is identified, so that the thickness of the plate section material can be identified subsequently; after identifying the plate material in the scrap steel image, segmenting the edge of the identified plate material to eliminate the rest interference information in the image, only segmenting the edge of the plate material to generate a plate material edge image only with the plate material edge, wherein the plate material edge image can be a binary image or an image display form which can obviously distinguish the plate material edge from other parts in the image in any other form; and calculating the thickness of the plate profile according to the obtained edge image of the plate profile so as to judge the plate profile according to the thickness of the plate profile. According to the method, the plate material edge image is firstly segmented, then the plate material thickness is calculated based on the segmented plate material edge image, the characteristics are extracted and the thickness is calculated to be divided into two steps, so that the plate material edge characteristics are clear, the identification is convenient, meanwhile, the calculation accuracy of the plate material thickness is ensured due to the clear plate material edge characteristics, and the accuracy of plate material grading is improved.
In the above embodiment, before identifying the slab in the scrap image, the method further includes: acquiring a scrap steel image; and adjusting the size of the scrap steel image to obtain the scrap steel image with the preset size.
In the embodiment, the scrap steel image is obtained firstly, so that the subsequent identification step is conveniently carried out according to the scrap steel image, the size of the obtained scrap steel image is adjusted, the size of the scrap steel image is adjusted to be the preset size, the scrap steel image during image identification at each time is processed to be the same size, the same distribution characteristics can be obtained during subsequent image processing, the influence of image size change on detection of the thickness of the plate type material according to the scrap steel image is avoided, the condition that the thickness of the plate type material in the image is changed due to adjustment of the size of the image is avoided, and the accuracy of scrap steel image identification is improved.
In the above embodiment, the step of adjusting the size of the scrap image to obtain a scrap image with a preset size specifically includes: when the size of the scrap steel image is larger than the preset size, cutting the scrap steel image to the preset size; and when the size of the scrap steel image is smaller than the preset size, filling the scrap steel image to the preset size.
In this embodiment, the size of the scrap image is adjusted to a preset size by comparing and adjusting the size of the scrap image with the preset size, specifically, when the size of the scrap image is larger than the preset size, the scrap image is cut, the size of the scrap image is adjusted to the preset size, when the size of the scrap image is smaller than the preset size, the scrap image is filled, and the size of the scrap image is filled to the preset size, so that the size of the scrap image is adjusted to a uniform size, so that the same distribution of features can be obtained in subsequent image recognition, compared with a scaling method, the size of the image is adjusted.
In the above embodiment, the step of performing edge segmentation on the identified sheet material to generate an edge image of the sheet material specifically includes: and performing edge segmentation on the plate material in the scrap steel image according to the semantic segmentation model to generate a plate material edge image.
In the embodiment, the edge of the plate material in the scrap steel image is segmented through the semantic segmentation model to generate a plate material edge image, so that the thickness of the plate material is calculated based on the plate material edge image, meanwhile, the semantic segmentation model can well combine global and local features in the scrap steel image to segment the plate material edge of the main body in the scrap steel image, the segmentation result is more accurate, interference information in the image can be eliminated, and the segmentation result is more accurate.
In the above embodiment, the step of calculating the thickness of the sheet material according to the edge image of the sheet material specifically includes: calculating the perimeter of the plate type material edge in the plate type material edge image and the area of the plate type material edge; and obtaining the thickness of the plate profile according to the perimeter and the area of the edge of the plate profile.
In this embodiment, the perimeter and the area of the sheet edge in the sheet edge image are calculated to calculate the thickness of the sheet according to the calculated perimeter and area results, so that the sheet can be classified according to the calculated thickness of the sheet. The specific thickness calculation method can be understood as moving the pixel point position in the picture to make the shape of the pixel point position be a regular shape, and since the perimeter and the area of the whole graph are not changed in the pixel point moving process, the calculation of the thickness becomes simple after the pixel point position is adjusted to obtain a regular graph (the calculation is only convenient to understand and does not represent the actual calculation process).
In the above embodiment, the step of calculating the perimeter of the panel edge and the area of the panel edge in the panel edge image specifically includes: calculating the perimeter of the plate molding material edge in the plate molding material edge image through 8-field algorithm; and calculating the area of the plate type material edge in the plate type material edge image by a threshold value method.
In the embodiment, the perimeter and the thickness of the edge of the plate type material are respectively calculated through an 8-field algorithm and a threshold value method, so that the thickness of the plate type material is calculated according to the relationship between the perimeter and the area of the edge of the plate type material, and the plate type material is judged according to the thickness of the plate type material.
In a second aspect, an embodiment of the present invention provides a system 200 for identifying a plate profile thickness in a scrap image, as shown in fig. 2, comprising: the identification module 202 is used for identifying the plate type materials in the scrap steel image; a segmentation module 204, configured to perform edge segmentation on the identified plate material to generate a plate material edge image; a calculating module 206, configured to calculate the thickness of the plate material according to the plate material edge image.
The system 200 for identifying the thickness of the plate material in the scrap steel image provided by the embodiment of the invention comprises a feature identification module 202, a segmentation module 204 and a calculation module 206. The identification module 202 is used for identifying the plate type materials in the scrap steel image; the segmentation module 204 is configured to perform edge segmentation on the identified plate material to generate a plate material edge image; the calculation module 206 is configured to calculate the thickness of the sheet according to the sheet edge image. Meanwhile, according to the system for identifying the plate type thickness in the steel scrap image provided by the embodiment of the invention, since the system is used for implementing the steps of the method for identifying the plate type thickness in the steel scrap image provided by the first aspect of the invention, the system for identifying the plate type thickness in the steel scrap image has all technical effects of the method for identifying the plate type thickness in the steel scrap image, and details are not repeated here.
In the above embodiment, the system for identifying the plate profile thickness in the scrap image further comprises: and the preprocessing module is used for acquiring the scrap steel image and adjusting the size of the scrap steel image to obtain the scrap steel image with the preset size.
In the embodiment, the scrap steel image is obtained firstly, so that the subsequent identification step is conveniently carried out according to the scrap steel image, the size of the obtained scrap steel image is adjusted, the size of the scrap steel image is adjusted to be the preset size, the scrap steel image during image identification at each time is processed to be the same size, the same distribution characteristics can be obtained during subsequent image processing, the influence of image size change on detection of the thickness of the plate type material according to the scrap steel image is avoided, the condition that the thickness of the plate type material in the image is changed due to adjustment of the size of the image is avoided, and the accuracy of scrap steel image identification is improved.
In the above embodiment, the preprocessing module is specifically configured to: when the size of the scrap steel image is larger than the preset size, cutting the scrap steel image to the preset size; and when the size of the scrap steel image is smaller than the preset size, filling the scrap steel image to the preset size.
In this embodiment, the size of the scrap image is adjusted to a preset size by comparing and adjusting the size of the scrap image with the preset size, specifically, when the size of the scrap image is larger than the preset size, the scrap image is cut, the size of the scrap image is adjusted to the preset size, when the size of the scrap image is smaller than the preset size, the scrap image is filled, and the size of the scrap image is filled to the preset size, so that the size of the scrap image is adjusted to a uniform size, so that the same distribution of features can be obtained in subsequent image recognition, compared with a scaling method, the size of the image is adjusted.
In the above embodiment, the segmentation module is specifically configured to: and performing edge segmentation on the plate material in the scrap steel image according to the semantic segmentation model to generate a plate material edge image.
In the embodiment, the edge of the plate material in the scrap steel image is segmented through the semantic segmentation model to generate a plate material edge image, so that the thickness of the plate material is calculated based on the plate material edge image, meanwhile, the semantic segmentation model can well combine global and local features in the scrap steel image to segment the plate material edge of the main body in the scrap steel image, the segmentation result is more accurate, interference information in the image can be eliminated, and the segmentation result is more accurate.
In the above embodiment, the calculation module is specifically configured to: calculating the perimeter of the plate molding material edge and the area of the plate molding material edge in the plate molding material edge image; and obtaining the thickness of the plate profile according to the perimeter and the area of the edge of the plate profile.
In this embodiment, the perimeter and the area of the sheet edge in the sheet edge image are calculated to calculate the thickness of the sheet according to the calculated perimeter and area results, so that the sheet can be classified according to the calculated thickness of the sheet. The specific thickness calculation method can be understood as moving the pixel point position in the picture to make the shape of the pixel point position be a regular shape, and since the perimeter and the area of the whole graph are not changed in the pixel point moving process, the calculation of the thickness becomes simple after the pixel point position is adjusted to obtain a regular graph (the calculation is only convenient to understand and does not represent the actual calculation process).
In the above embodiment, the calculation module is further specifically configured to: calculating the perimeter of the plate molding material edge in the plate molding material edge image through 8-field algorithm; and calculating the area of the plate profile edge in the plate profile edge image by using a threshold value method.
In the embodiment, the perimeter and the thickness of the edge of the plate type material are respectively calculated through an 8-field algorithm and a threshold value method, so that the thickness of the plate type material is calculated according to the relationship between the perimeter and the area of the edge of the plate type material, and the plate type material is judged according to the thickness of the plate type material.
A third aspect embodiment of the present invention provides a system 300 for identifying plate profile thickness in a scrap image, as shown in fig. 3, comprising: a memory 302, a processor 304 and a program stored on the memory 302 and executable on the processor 304, the program when executed by the processor 304 implementing the steps defined in the method of identifying plate thickness in a scrap image of any of the embodiments described above.
The system for identifying the plate section thickness in the steel scrap image comprises a memory, a processor and a program stored on the memory and capable of running on the processor, wherein the program realizes the steps defined by any one of the above methods for identifying the plate section thickness in the steel scrap image when being executed by the processor. Meanwhile, the system for identifying the thickness of the plate type material in the steel scrap image can realize the steps limited by any method for identifying the thickness of the plate type material in the steel scrap image, so that the system for identifying the thickness of the plate type material in the steel scrap image provided by the embodiment has all the beneficial effects of the method for identifying the thickness of the plate type material in the steel scrap image provided by any embodiment.
A fourth aspect embodiment of the present invention provides a readable storage medium having stored thereon a program and/or instructions which, when executed by a processor, carry out the steps of the method of identifying plate section thickness in a scrap image in any of the above embodiments.
According to the readable storage medium provided by the embodiment of the present invention, since the program and/or the instructions stored thereon can be executed by the processor to implement the steps of the method for identifying plate type thickness in a steel scrap image in any of the above embodiments, all beneficial technical effects of the method for identifying plate type thickness in a steel scrap image are provided, and are not described herein again.
The fifth embodiment of the invention provides an intelligent steel scrap grading system, which comprises a system for identifying the thickness of a plate material in a steel scrap image in the embodiment; or a readable storage medium as in the above embodiments.
The intelligent steel scrap grading system provided according to the embodiment of the invention is provided because it includes the system for identifying the thickness of the plate profile in the steel scrap image as in the above embodiment or the readable storage medium as in the above embodiment. Therefore, the intelligent scrap steel grading system has all the technical effects of the system for identifying the thickness of the plate section in the scrap steel image or the readable storage medium, and the details are not repeated herein.
The method for identifying the plate profile thickness in the scrap image provided by the present application will be further described with reference to another embodiment.
The method for identifying the thickness of the plate profile in the scrap image according to this embodiment is shown in fig. 4, and includes:
s402, preprocessing the image.
Specifically, the requirement of a semantic segmentation model is met by preprocessing a focusing image of the ball machine. After the vehicle enters the discharging point, as shown in fig. 5, a focusing image is shot through the cooperation of a gun-ball machine, the image can be subjected to material type classification through a single instance segmentation model, as shown in fig. 6, plate type materials in the focusing image are obtained, and therefore thickness identification is carried out on the plate type materials according to the classification result. Because the method is an image-based recognition method, and the thickness of an object in an image has a problem of 'near-large-far-small', a resize operation is not performed when a single body in the image is processed, namely, when the size of the image is adjusted, the image is not zoomed, but crop and padding operations are uniformly adopted, namely, the image is cropped and padded, padding (padding) is performed on the image smaller than the standard size, namely, one block is supplemented to the image on the original basis, crop (cropping) is performed on the image larger than the standard size, and a part larger than the standard size is cropped. Thus, under the condition that the distance between the camera and the object is the same and the focal distance is the same, the thickness scales displayed by different images are the same.
And S404, semantic segmentation.
Specifically, edge segmentation is performed on the preprocessed image, and a semantic segmentation model in a neural network is used. The semantic segmentation model can combine global and local features in the image, as shown in fig. 7, segment the sheet edge of the main body in the image, and exclude interference information.
And S406, post-processing.
Specifically, after the edge of the plate material is identified, the plate material is only a binary image and cannot be converted into specific thickness information, and the thickness of the edge of the binary image is calculated through post-processing operation to obtain a specific value. And in the post-processing operation, the perimeter of the edge image is obtained through an 8-field algorithm, the area of the thickness image is obtained according to a threshold value method, and the thickness is calculated according to the relationship between the perimeter and the area.
According to the method for identifying the plate section thickness in the scrap steel image, which is provided by the embodiment, the relations among image preprocessing, semantic segmentation and post-processing are serial.
The semantic segmentation and post-processing stages in the present application were verified through experiments.
Evaluation indexes in the segmentation stage are as follows:
pixel Accuracy (PA, pixel Accuracy): marking the percentage of correct pixels to the total pixels;
recall (Recall): the proportion of the sample with the predicted value of 1 and the real value of 1 in all the samples with the real value of 1;
and in the post-processing stage, predicting the number of the monomers with the thickness of +/-2 mm = the real thickness, namely judging the correct number of the monomers, wherein the evaluation indexes are as follows:
accuracy (Accuracy): determining the correct number of monomers/total number of monomers
This experiment verified a total of 300 monomers.
The results of the experiment are shown in table 1:
TABLE 1
PA Recall
Monomer with diameter of more than 6mm 99.40％ 98.14％
Monomer with diameter of less than 6mm 95.10％ 90.24％
Generally, before entering the neural network, a certain preprocessing operation is performed, wherein resize is the most widely used, and the neural network is guaranteed to obtain the same distributed features by processing the image to the same scale. When thickness recognition is performed, resize adjusts the scale of the image, and similarly, the scale of the thickness also changes, and the result is inaccurate. The crop and padding modes adopted by the people can not only adjust the images to the same scale, but also ensure that the thickness scales of two different images are the same. Meanwhile, because the difference between the appearances of the thick plate type material and the thin plate type material is not large, the only difference is the thickness of the edge, and the part of the characteristics are not obvious in the whole image, and the effective characteristics cannot be extracted by a neural network. Therefore, the method has the advantages that the characteristic extraction and the thickness calculation are separately calculated, the characteristics are clear, the identification is accurate, and the interpretability is strong.
In this specification, the term "plurality" means two or more unless explicitly defined otherwise. The terms "mounted," "connected," "fixed," and the like are to be construed broadly, and for example, "connected" may be a fixed connection, a removable connection, or an integral connection; "coupled" may be direct or indirect through an intermediary. The specific meanings of the above terms in the present invention can be understood by those skilled in the art according to specific situations.
In the description of the present specification, the description of the terms "one embodiment," "some embodiments," or the like, means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above do not necessarily refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
The above is only a preferred embodiment of the present invention, and is not intended to limit the present invention, and various modifications and changes will occur to those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims (10)

1. A method for identifying the thickness of a plate material in a scrap steel image is used for an intelligent scrap steel grading system and is characterized by comprising the following steps:
identifying plate type materials in the scrap steel image;
performing edge segmentation on the identified plate type material to generate a plate type material edge image;
and calculating the thickness of the plate type material according to the edge image of the plate type material.
2. The method of identifying plate profile thickness in a scrap image in accordance with claim 1, further comprising, prior to said identifying plate profile in a scrap image:
acquiring a scrap steel image;
and adjusting the size of the scrap steel image to obtain the scrap steel image with a preset size.
3. The method for identifying the plate profile thickness in the scrap image according to claim 2, wherein the step of adjusting the size of the scrap image to obtain the scrap image with the preset size comprises the following steps:
when the size of the scrap steel image is larger than a preset size, cutting the scrap steel image to the preset size;
and when the size of the scrap steel image is smaller than a preset size, filling the scrap steel image to the preset size.
4. The method of claim 1, wherein the step of performing edge segmentation on the identified plate profile to generate a plate profile edge image comprises:
and performing edge segmentation on the plate material in the scrap steel image according to a semantic segmentation model to generate a plate material edge image.
5. The method of claim 1, wherein the step of calculating the thickness of the plate profile according to the plate profile edge image comprises:
calculating the perimeter of the plate molding material edge and the area of the plate molding material edge in the plate molding material edge image;
and obtaining the thickness of the plate type material according to the perimeter and the area of the edge of the plate type material.
6. The method of claim 4, wherein the step of calculating the perimeter of the slab edge and the area of the slab edge in the slab edge image comprises:
calculating the perimeter of the plate profile edge in the profile edge image through an 8-field algorithm;
and calculating the area of the plate profile edge in the plate profile edge image by a threshold value method.
7. A system for identifying plate profile thickness in scrap image comprising:
the identification module is used for identifying the plate type materials in the scrap steel image;
the segmentation module is used for performing edge segmentation on the identified plate type material to generate a plate type material edge image;
and the calculating module is used for calculating the thickness of the plate type material according to the edge image of the plate type material.
8. A system for identifying plate profile thickness in a scrap image, comprising a memory and a processor, the memory storing a program or instructions executable on the processor, the program or instructions when executed by the processor implementing the steps of the method of identifying plate profile thickness in a scrap image as claimed in any one of claims 1 to 6.
9. A readable storage medium, on which a program and/or instructions are stored, which when executed by a processor, carry out the steps of the method of identifying plate profile thickness in a scrap image according to any one of claims 1 to 6.
10. An intelligent scrap steel grading system comprising the system for identifying sheet profile thickness in scrap steel images according to claim 7 or 8; or
The readable storage medium of claim 9.
CN202210766116.0A 2022-07-01 2022-07-01 Method and system for identifying plate material thickness in scrap steel image Pending CN115222678A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210766116.0A CN115222678A (en) 2022-07-01 2022-07-01 Method and system for identifying plate material thickness in scrap steel image

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210766116.0A CN115222678A (en) 2022-07-01 2022-07-01 Method and system for identifying plate material thickness in scrap steel image

Publications (1)

Publication Number Publication Date
CN115222678A true CN115222678A (en) 2022-10-21

Family

ID=83609081

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210766116.0A Pending CN115222678A (en) 2022-07-01 2022-07-01 Method and system for identifying plate material thickness in scrap steel image

Country Status (1)

Country Link
CN (1) CN115222678A (en)

Similar Documents

Publication Publication Date Title
CN115082467B (en) Building material welding surface defect detection method based on computer vision
CN113724231B (en) Industrial defect detection method based on semantic segmentation and target detection fusion model
US7373018B2 (en) Method and system for object recognition using fractal maps
US7257267B2 (en) Method and system for image segmentation
CN110544231B (en) Lithium battery electrode surface defect detection method based on background standardization and centralized compensation algorithm
CN116559183B (en) Method and system for improving defect judging efficiency
CN111815555A (en) Metal additive manufacturing image detection method and device combining anti-neural network with local binary
CN116777907A (en) Sheet metal part quality detection method
CN115719332A (en) Welding quality detection method
CN115953373B (en) Glass defect detection method, device, electronic equipment and storage medium
CN117351001B (en) Surface defect identification method for regenerated aluminum alloy template
CN118279304B (en) Abnormal recognition method, device and medium for special-shaped metal piece based on image processing
CN117392097A (en) Additive manufacturing process defect detection method and system based on improved YOLOv8 algorithm
CN115229355A (en) Laser cutting method for high-precision stamping forging
CN118052823A (en) Titanium plate welding spot defect detection method based on image processing
CN114004858A (en) Method and device for identifying aviation cable surface code based on machine vision
CN116258838B (en) Intelligent visual guiding method for duct piece mold clamping system
CN115222678A (en) Method and system for identifying plate material thickness in scrap steel image
CN116363136B (en) On-line screening method and system for automatic production of motor vehicle parts
CN115578390B (en) Welding control method for deaerator
CN114943843B (en) Welding defect detection method based on shape sensing
CN114841971A (en) Method, system and device for detecting pit defects of metal block
US10482314B2 (en) Automatic calculation for ploidy classification
CN117152447B (en) Intelligent management method and system for punching die
CN117132580B (en) Method and system for measuring solidification structure based on dendrite macroscopic inspection image

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination