CN113570599A - Image processing-based method and system for evaluating quality of solid wood particle board material - Google Patents

Image processing-based method and system for evaluating quality of solid wood particle board material Download PDF

Info

Publication number
CN113570599A
CN113570599A CN202111104111.3A CN202111104111A CN113570599A CN 113570599 A CN113570599 A CN 113570599A CN 202111104111 A CN202111104111 A CN 202111104111A CN 113570599 A CN113570599 A CN 113570599A
Authority
CN
China
Prior art keywords
image
wood
tone
images
holes
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.)
Granted
Application number
CN202111104111.3A
Other languages
Chinese (zh)
Other versions
CN113570599B (en
Inventor
段宝光
陈玉军
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.)
Nantong Beilei Exhibition Cabinet Manufacturing Co ltd
Original Assignee
Jiangsu Lvquan Decoration Engineering 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 Jiangsu Lvquan Decoration Engineering Co ltd filed Critical Jiangsu Lvquan Decoration Engineering Co ltd
Priority to CN202111104111.3A priority Critical patent/CN113570599B/en
Publication of CN113570599A publication Critical patent/CN113570599A/en
Application granted granted Critical
Publication of CN113570599B publication Critical patent/CN113570599B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

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
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
    • G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/90—Determination of colour characteristics
    • 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
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30108—Industrial image inspection
    • G06T2207/30161—Wood; Lumber

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Human Resources & Organizations (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Development Economics (AREA)
  • Quality & Reliability (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Educational Administration (AREA)
  • Economics (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Strategic Management (AREA)
  • Game Theory and Decision Science (AREA)
  • Marketing (AREA)
  • Operations Research (AREA)
  • Tourism & Hospitality (AREA)
  • General Business, Economics & Management (AREA)
  • Image Analysis (AREA)

Abstract

The invention discloses a method and a system for evaluating the quality of a solid wood particle board based on image processing, wherein the method comprises the following steps: carrying out semantic segmentation on the collected wood board section image to obtain a wood board section image; carrying out image conversion to obtain a wood section HSV image, extracting a wood board tone channel image, and calculating a tone consistency index; carrying out gray level conversion to obtain a gray level image, and carrying out level division on the gray level image to obtain different level gray level images; performing threshold segmentation and reverse threshold processing to obtain reverse threshold images of different levels, and taking the reverse threshold image with a pixel value of 1 as a hole; calculating the hole influence values of the reverse threshold images of different levels, obtaining a board quality evaluation value according to the hole influence values and the tone consistency indexes, and evaluating the quality of the particle board by using the quality evaluation value. Through image processing and layered processing, the quality evaluation of the particle board is realized, the subjectivity and the low efficiency of artificial detection are avoided, and the detection efficiency and the accuracy are improved.

Description

Image processing-based method and system for evaluating quality of solid wood particle board material
Technical Field
The application relates to the field of artificial intelligence, in particular to a method and a system for evaluating the quality of a solid wood particle board material based on image processing.
Background
At present, the quality of the obtained particle board is mainly evaluated through manual detection, the manual detection method is easily interfered by various factors such as the experience of detection personnel, the state of the personnel, the observation angle and the like, and the detection process has subjectivity and low detection efficiency due to low precision.
Disclosure of Invention
Aiming at the technical problems, the invention provides a method and a system for evaluating the quality of a solid wood particle board based on image processing.
In a first aspect, the present disclosure provides an image processing-based method for evaluating a quality of a solid wood particle board, including:
carrying out semantic segmentation on the collected wood board section image to obtain a wood board section image;
converting the wood board section image to obtain a wood section HSV image, extracting a wood board tone channel image, calculating tone variance of the tone channel image, and taking the tone variance as a tone consistency index;
carrying out gray level conversion on the wood board section image to obtain a gray level image, and carrying out level division on the gray level image to obtain different level gray level images;
performing threshold segmentation on the gray level images of different levels to obtain threshold segmentation images, performing reverse threshold processing on the threshold segmentation images to obtain reverse threshold images of different levels, and taking the image with a pixel value of 1 in the reverse threshold images as a hole;
calculating the hole influence values of the reverse threshold images of different levels, obtaining a board quality evaluation value according to the hole influence values and the tone consistency indexes, and evaluating the quality of the particle board by using the quality evaluation value;
wherein, the calculation model of the wood board quality evaluation value is as follows:
Figure 100002_DEST_PATH_IMAGE002
in the formula
Figure 100002_DEST_PATH_IMAGE004
Refers to the tone uniformity of a tone channel image of a wood boardAn evaluation value of the evaluation value is calculated,
Figure 100002_DEST_PATH_IMAGE006
is as follows
Figure 100002_DEST_PATH_IMAGE008
Influence values of holes in the image of the reverse threshold value of each level;
the holes comprise large holes and small holes; the calculation model of the influence value of the holes in the reverse threshold value image at each level is as follows:
Figure 100002_DEST_PATH_IMAGE010
in the formula
Figure 100002_DEST_PATH_IMAGE012
Is as follows
Figure 42986DEST_PATH_IMAGE008
The total number of small holes in the reverse threshold image of each level,
Figure 100002_DEST_PATH_IMAGE014
is as follows
Figure 184118DEST_PATH_IMAGE008
The total number of large holes in the individual level inverse threshold image,
Figure 100002_DEST_PATH_IMAGE016
first finger
Figure 100002_DEST_PATH_IMAGE018
The area of each large hole is larger than that of the corresponding hole,
Figure 100002_DEST_PATH_IMAGE020
first finger
Figure 917981DEST_PATH_IMAGE018
A large hole and the first
Figure 100002_DEST_PATH_IMAGE022
The shortest distance between the big holes,
Figure 100002_DEST_PATH_IMAGE024
the average area of all the large holes, the large hole is a connected domain with the area larger than the average area of all the connected domains, the small hole is a connected domain with the area not larger than the average area of all the connected domains,
Figure 100002_DEST_PATH_IMAGE026
wherein
Figure 100002_DEST_PATH_IMAGE028
Refers to the longest distance between large cavities.
Further, the method for performing hierarchical division on the gray-scale image comprises the following steps:
calculating the variance value of the gray level image, then moving downwards pixel by pixel from the upper part of the image by a straight line parallel to the horizontal direction, and respectively calculating the in-class variances at two sides of the straight line after moving one pixel each time
Figure 100002_DEST_PATH_IMAGE030
And
Figure 100002_DEST_PATH_IMAGE032
；
Figure 100002_DEST_PATH_IMAGE034
，
Figure 100002_DEST_PATH_IMAGE036
and representing the inter-class variance, when the inter-class variance is maximum, minimizing the intra-class variance, taking the straight line as a dividing line, completing the hierarchical division of the upper image and the lower image, and dividing the lower image again until the gray image is divided, thereby finally obtaining the gray images with different levels.
Further, the calculation model of the hue consistency evaluation value is as follows:
Figure 100002_DEST_PATH_IMAGE038
in the formula
Figure 100002_DEST_PATH_IMAGE040
A reference numeral is given to each of the pixels,
Figure 100002_DEST_PATH_IMAGE042
represents the total number of pixels and the total number of pixels,
Figure 100002_DEST_PATH_IMAGE044
represents the first
Figure 404807DEST_PATH_IMAGE040
The value of each of the pixels is calculated,
Figure 100002_DEST_PATH_IMAGE046
representing the pixel mean.
In a second aspect, the invention provides an image processing-based evaluation system for quality of solid wood particle board, comprising:
the wood slice image acquisition unit is used for performing semantic segmentation on the acquired wood board section image to obtain a wood board section image;
the tone consistency calculating unit is used for converting the wood board section image to obtain a wood section HSV image, extracting a wood board tone channel image, calculating tone variance of the tone channel image and taking the tone variance as a tone consistency index;
the gray image level division unit is used for carrying out gray level conversion on the wood board section image to obtain a gray image and carrying out level division on the gray image to obtain different level gray images;
the image reverse threshold processing unit is used for carrying out threshold segmentation on the gray level images of different levels to obtain threshold segmentation images, carrying out reverse threshold processing on the threshold segmentation images to obtain reverse threshold images of different levels, and taking the image with the pixel value of 1 in the reverse threshold images as a hole;
the quality evaluation value calculation unit is used for calculating the hole influence values of the reverse threshold images of different layers, obtaining a board quality evaluation value according to the hole influence values and the tone consistency indexes, and evaluating the quality of the particle board by using the quality evaluation value; wherein, the calculation model of the wood board quality evaluation value is as follows:
Figure 100002_DEST_PATH_IMAGE002A
in the formula
Figure 460838DEST_PATH_IMAGE004
Refers to the tone consistency evaluation value of the wood tone channel image,
Figure 546606DEST_PATH_IMAGE006
is as follows
Figure 351619DEST_PATH_IMAGE008
Influence values of holes in the image of the reverse threshold value of each level;
the holes comprise large holes and small holes; the calculation model of the influence value of the holes in the reverse threshold value image at each level is as follows:
Figure 100002_DEST_PATH_IMAGE010A
in the formula
Figure 484048DEST_PATH_IMAGE012
Is as follows
Figure 374512DEST_PATH_IMAGE008
The total number of small holes in the reverse threshold image of each level,
Figure 123026DEST_PATH_IMAGE014
is as follows
Figure 798858DEST_PATH_IMAGE008
The total number of large holes in the individual level inverse threshold image,
Figure 775558DEST_PATH_IMAGE016
first finger
Figure 28685DEST_PATH_IMAGE018
The area of each large hole is larger than that of the corresponding hole,
Figure 987413DEST_PATH_IMAGE020
first finger
Figure 642386DEST_PATH_IMAGE018
A large hole and the first
Figure 647775DEST_PATH_IMAGE022
The shortest distance between the big holes refers to the average area of all the big holes, the big holes refers to the connected domain with the area larger than the average area of all the connected domains, the small holes refers to the connected domain with the area not larger than the average area of all the connected domains,
Figure 653777DEST_PATH_IMAGE026
wherein
Figure 806410DEST_PATH_IMAGE028
Refers to the longest distance between large cavities.
Compared with the traditional technical scheme, the beneficial effect brought by one aspect of the disclosure is as follows:
1. the quality evaluation is carried out on the solid wood particle board based on the image processing, and the detection efficiency and the detection precision can be improved.
2. Holes in the solid wood particle board are classified, the influence of the holes on the quality of the board is respectively considered, and a more accurate quality detection result can be effectively obtained.
Drawings
Embodiments herein will be described in more detail, by way of example only, with reference to the accompanying drawings, in which:
fig. 1 is a block diagram of a method for evaluating the quality of a solid wood particle board based on image processing.
Fig. 2 is a gray scale view of particle board wood in an example of the invention.
Fig. 3 is a schematic diagram of gradation division of a grayscale image in an embodiment of the present invention.
FIG. 4 is a flowchart of evaluation values of the quality of the wooden board according to an embodiment of the present invention.
FIG. 5 is a schematic diagram of the area and path of the large hole in the embodiment of the present invention.
Fig. 6 is a block diagram of a system for evaluating the quality of a solid wood particle board based on image processing according to the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
In the following description, for purposes of explanation and not limitation, specific details are set forth, such as particular system structures, techniques, etc. in order to provide a thorough understanding of the embodiments of the present application. It will be apparent, however, to one skilled in the art that the present application may be practiced in other embodiments that depart from these specific details. In other instances, detailed descriptions of well-known systems, devices, circuits, and methods are omitted so as not to obscure the description of the present application with unnecessary detail.
The terms "first", "second" and "first" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature; in the description of the present embodiment, "a plurality" means two or more unless otherwise specified.
Fig. 1 is a block diagram of a method for evaluating a quality of a solid wood particle board based on image processing according to this embodiment, and the method for evaluating a quality of a solid wood particle board based on image processing shown in fig. 1 includes the following steps:
s101, performing semantic segmentation on the collected wood board section image to obtain a wood board section image, wherein the wood board section image specifically comprises the following contents;
the purpose of the steps is as follows: the target and the background of the wood board section image are segmented through semantic segmentation to obtain a wood board section image, and the interference of irrelevant factors such as the background and the like to the image processing is eliminated.
In this embodiment, semantic segmentation is performed through DNN, the used data set is a wood board section image data set acquired in an overlooking manner, and pixels to be segmented are divided into two types, that is, the label labeling process corresponding to the training set is a single-channel semantic label, the label of a corresponding position pixel belonging to the background class is 0, and the label of a corresponding position pixel belonging to the wood board is 1.
And S102, converting the wood board section image to obtain a wood section HSV image, extracting to obtain a wood board tone channel image, calculating tone variance of the tone channel image, and taking the tone variance as a tone consistency index.
The step aims to obtain a wood tone channel image, provide basis for subsequent tone consistency index calculation, and consider the tone consistency index of the particle board
Figure 456834DEST_PATH_IMAGE004
The influence on the quality of the plate, the color of the good particle plate in the embodiment is more consistent, and the variance of the tone image
Figure 974403DEST_PATH_IMAGE004
The larger the plate, the worse the plate quality; in contrast, tonal image variance
Figure 589406DEST_PATH_IMAGE004
The smaller the sheet quality the better.
The wood board section image is an RGB image, but the RGB color space is far less friendly than an HSV color space, different colors can be distinguished more easily in the HSV color space, the wood slice image is converted from the RGB to the HSV color space, and the obtained HSV image has three channels: and extracting the hue channel image, the saturation channel and the brightness channel.
Calculating variance of tone image
Figure 358778DEST_PATH_IMAGE004
The variance of the tone image is used as a tone consistency index,
Figure 100002_DEST_PATH_IMAGE048
wherein each pixel is represented by a reference numeral,
Figure 440873DEST_PATH_IMAGE042
represents the total number of pixels and the total number of pixels,
Figure DEST_PATH_IMAGE050
representing the value of the first pixel or pixels,
Figure 457239DEST_PATH_IMAGE046
representing the pixel mean.
Step S103, carrying out gray level transformation on the wood board section image to obtain a gray level image, and carrying out level division on the gray level image to obtain different level gray level images, wherein the method specifically comprises the following steps:
the step aims to perform longitudinal hierarchical division on the gray level image after performing gray level conversion on the particle board section image, and divide the gray level image into different levels so as to calculate the board quality evaluation value in each level subsequently.
Fig. 2 shows a gray scale image of the particle board wood in this embodiment, as shown in fig. 2, the upper and lower layers of the particle board in this embodiment use fine wood fibers, and the middle interlayer uses long wood fibers, so that the particles of the upper and lower layers are smaller, the holes are smaller and smoother, the particles of the middle portion are larger, and a distinct layering is formed between the upper and lower layers of the gray scale image of the wood board.
Firstly, obtaining a gray level image, carrying out gray level conversion on a board slice image in an RGB format,
Figure DEST_PATH_IMAGE052
wherein R, G, B represents the red channel and green channel of the RGB image respectivelyPixel values for the channel and blue channel.
In the embodiment, the gray level image has three levels, so that the gray level image level distribution is conveniently and quickly obtained, a method of gradually dividing the whole gray level image from top to bottom to obtain a plurality of levels is not adopted, the first dividing line is directly found from top to bottom, and then the second dividing line is found from bottom to top.
In this embodiment, the inter-class variance + the intra-class variance = the image variance, and fig. 3 shows a schematic diagram of hierarchical division of the grayscale image in this embodiment, when two dividing lines are located at the positions shown in fig. 3, the intra-class variance is the smallest, that is, the inter-class variance is the largest, and the hierarchical division is more obvious at this time.
First, find the first dividing line
Figure DEST_PATH_IMAGE054
Firstly, the variance value of the gray level image is calculated
Figure DEST_PATH_IMAGE056
. Then, a straight line parallel to the horizontal direction is used for moving downwards pixel by pixel from the upper part of the image, and after one pixel is moved each time, the variance values at two sides of the straight line are respectively calculated
Figure 690031DEST_PATH_IMAGE030
And
Figure 263095DEST_PATH_IMAGE032
。
by using
Figure 606220DEST_PATH_IMAGE036
To express the total variance of the image
Figure 340958DEST_PATH_IMAGE056
And
Figure 274760DEST_PATH_IMAGE030
、
Figure 900783DEST_PATH_IMAGE032
the difference value of (a) to (b),
Figure 114726DEST_PATH_IMAGE034
Figure DEST_PATH_IMAGE058
representing the intra-class variance on both sides of the line,
Figure 128688DEST_PATH_IMAGE056
the variance of the entire image is represented,
Figure 428082DEST_PATH_IMAGE036
representing the inter-class variance, when the inter-class variance is maximum, the intra-class variance is minimum, the variance between the same classes is minimum, the similarity between the pixels at two sides of the straight line is maximum, and the straight line is the first dividing line
Figure 736091DEST_PATH_IMAGE054
。
Then searching for a second dividing line
Figure DEST_PATH_IMAGE060
Second dividing line
Figure 257071DEST_PATH_IMAGE060
Acquisition process and first split line
Figure 989404DEST_PATH_IMAGE054
The acquisition process is similar, a straight line parallel to the horizontal direction is used for moving from the lower part of the image pixel by pixel to the upper part, and after one pixel is moved each time, the sum of the variance values of the two sides of the straight line is respectively calculated
Figure 25362DEST_PATH_IMAGE032
Up to inter-class variance
Figure DEST_PATH_IMAGE062
At the maximum, the number of the first,
Figure DEST_PATH_IMAGE064
at this time, the straight lineIs a second dividing line
Figure 521415DEST_PATH_IMAGE060
At this time, the first dividing line
Figure 975530DEST_PATH_IMAGE054
And a second dividing line
Figure 613185DEST_PATH_IMAGE060
The grayscale image is divided into three levels.
And S104, performing threshold segmentation on the gray level images of different levels to obtain threshold segmentation images, performing reverse threshold processing on the threshold segmentation images to obtain reverse threshold images of different levels, and taking the image with the pixel value of 1 in the reverse threshold images as a hole.
The purpose of this step is to obtain a 1 connected domain through an inverse threshold image for all levels of gray scale images. In this embodiment, the gray level image is subjected to threshold segmentation by the Otsu method to obtain a threshold imageRThen by
Figure DEST_PATH_IMAGE066
And obtaining an inverse threshold value image, wherein the pixel value in the inverse threshold value image is 1 to represent the hole.
And S105, calculating the hole influence values of the reverse threshold images of different layers, and obtaining the board quality evaluation value according to the hole influence values and the tone consistency indexes. In this embodiment, the quality evaluation value of the whole particle board is obtained by calculating the hole influence value of each layer and finally integrating the hole influence values of all the layers and the tone consistency index of the whole wood board, and specifically includes the following contents:
fig. 4 shows a flowchart of the wood board quality evaluation value in the embodiment, and the calculation process of the wood board quality evaluation value will be described in detail with reference to fig. 4, which specifically includes the following steps:
calculating the hole influence value of each layer specifically comprises the following steps:
firstly, the area of a 1 connected domain is obtained through the total number of pixel points with the pixel value of 1 in a reverse threshold value imageS0, simultaneously obtaining the area of each connected domain, and obtaining the total number of the connected domains by an eight-neighborhood labeling algorithm
Figure 605281DEST_PATH_IMAGE040
Numbering the connected domains and obtaining the area average value of the connected domains
Figure DEST_PATH_IMAGE068
。
Secondly, the area of each connected domain
Figure DEST_PATH_IMAGE070
And area mean value
Figure DEST_PATH_IMAGE072
Compared with the prior art, the holes comprise large holes and small holes, and the area of the communicating region is larger than that of the small holes
Figure 504360DEST_PATH_IMAGE072
The connected region is marked as a large hole, and the area of the connected region is less than or equal to
Figure 348336DEST_PATH_IMAGE072
The connected domain of (a) is marked as a small hole. The quality influence of the small holes on the particle board is represented by the number of the small holes, and the number of the small holes is recorded as
Figure 547105DEST_PATH_IMAGE012
And is used to represent the influence value of small holes
Figure DEST_PATH_IMAGE074
I.e. by
Figure DEST_PATH_IMAGE076
。
Then calculating the shortest path between the big holes to obtain the shortest path vector and the hole size vector, and obtaining the big hole influence value
Figure DEST_PATH_IMAGE078
. Firstly, the number of the big holes from small to big is obtainedSequences and corresponding area sequences, and then replacing the numbering with the corresponding order results in a new numbered sequence, for example: the original numbering sequence is [3, 7, 9 ]]Using [1, 2, 3 ]]Replacing the original numbering sequence as a new numbering sequence.
The connected domain is represented by the center point of each connected domain, and the center point is (number, area of connected domain), for example: (3, 50) represents that the dot represents a connected component numbered 3, the area of which is 50.
In this embodiment, the shortest path between each connected domain point is calculated by Dijkstra algorithm, and a point with the number of 1 is used as a starting point, and the nodes are expanded layer by layer outward by taking the starting point as a center until the nodes are expanded to an end point. Therefore, a shortest path sequence value is obtained, and an area path diagram is obtained by combining the hole area sequence and the shortest path sequence.
FIG. 5 shows a schematic diagram of the area and path of large holes, as shown in FIG. 5, 1, 2, 3, 4, 5 represent connected domain numbers, box values represent the area of the holes, numbers between boxes represent the distance between different connected domains, and the area path diagram describes the size and distribution of the holes.
The number of large holes is denoted by k
Figure DEST_PATH_IMAGE080
Indicating the area of each hole by
Figure 419640DEST_PATH_IMAGE020
Is shown as
Figure 604633DEST_PATH_IMAGE018
A hole and the first
Figure 751450DEST_PATH_IMAGE022
The shortest distance between holes, e.g.
Figure DEST_PATH_IMAGE082
The area of the holes denoted by reference numeral 1,
Figure DEST_PATH_IMAGE084
the number of expression is 1 and the numberDistance between holes of 2.
In this embodiment, the sum of the products of the area and the distance of the corresponding hole is used to represent the influence value of the large hole
Figure DEST_PATH_IMAGE086
Then comprehensively calculating to obtain the hole influence value
Figure DEST_PATH_IMAGE088
The hole impact value of each level can be represented by the sum of the small hole impact value and the large hole impact value
Figure DEST_PATH_IMAGE090
Further, the influence values of all the hierarchical holes can be obtained,
Figure 416393DEST_PATH_IMAGE088
the larger the value the greater the effect of the holes and the poorer the quality of the particle board.
The normalization processing is performed on the hole influence value, and specifically includes the following steps:
since the value ranges of different evaluation values are not consistent, normalization is required to be performed respectively, and then the next calculation is performed.
The tone consistency index is evaluated by using a tone variance, the value range is within 0-1, and no additional operation is required; influence value of Small holes
Figure 55185DEST_PATH_IMAGE074
Expressed in terms of the number of small holes, can be divided by the total number of holes
Figure 653656DEST_PATH_IMAGE040
Carry out normalized mapping
Figure DEST_PATH_IMAGE092
。
To pair
Figure 858241DEST_PATH_IMAGE078
The normalization is performed by respectively comparing two parameters
Figure 274179DEST_PATH_IMAGE080
And
Figure 256392DEST_PATH_IMAGE020
the normalization is carried out, and the normalization is carried out,
Figure 48768DEST_PATH_IMAGE080
the normalization of (a) is realized by the average area s of the large holes in the whole layer, wherein s is obtained by dividing the total number of pixels of the large holes in the layer, namely the area of all the large holes by the number of the large holes; to pair
Figure 920909DEST_PATH_IMAGE020
The normalization method is to calculate the longest path between each large cavity under the level by the Bellman-Ford algorithm
Figure 632382DEST_PATH_IMAGE028
Then using the longest distance
Figure 714607DEST_PATH_IMAGE028
Is divided by
Figure DEST_PATH_IMAGE094
Obtaining the average longest distance among the large holes, and performing normalization treatment to show that:
Figure DEST_PATH_IMAGE096
wherein
Figure 438237DEST_PATH_IMAGE012
Represents the number of small holes in the film,
Figure 273207DEST_PATH_IMAGE040
representing the total number of holes.
Comprehensively obtaining a particle board quality evaluation value, which specifically comprises the following contents:
hole influence of each layer of particle board
Figure 762438DEST_PATH_IMAGE088
For the influence value of large holes per level
Figure 722173DEST_PATH_IMAGE074
And small pore influence value
Figure 121930DEST_PATH_IMAGE078
Is shown in the sum of (a) and (b),
Figure DEST_PATH_IMAGE098
。
hole influence values of all layers of the particle board
Figure DEST_PATH_IMAGE100
And, and
Figure DEST_PATH_IMAGE102
the sum of (1) is expressed by adding the color tone consistency index to the hole influence values of all the layers
Figure 673391DEST_PATH_IMAGE004
Obtaining the quality evaluation value of the particle board
Figure DEST_PATH_IMAGE104
The calculation formula is as follows:
Figure DEST_PATH_IMAGE106
，
Figure 590832DEST_PATH_IMAGE104
the smaller the value, the better the particle board quality.
The system of the embodiment of the invention is described below with reference to fig. 6, and fig. 6 shows a block diagram of an image processing-based evaluation system for the quality of solid wood particle board, and as shown in the figure, the image processing-based evaluation system for the quality of solid wood particle board comprises the following contents:
and the wood slice image acquisition unit 201 is used for performing semantic segmentation on the acquired wood board section image to obtain a wood board section image.
And the hue consistency calculating unit 202 is configured to convert the wood board section image to obtain a wood section HSV image, extract to obtain a wood board hue channel image, calculate a hue variance of the hue channel image, and use the hue variance as a hue consistency index.
And the gray image level division unit 203 is used for performing gray level conversion on the wood board section image to obtain a gray image and performing level division on the gray image to obtain different levels of gray images.
An image reverse threshold processing unit 204, configured to perform threshold segmentation on the grayscale images of different levels to obtain threshold segmentation images, and perform reverse threshold processing on the threshold segmentation images to obtain reverse threshold images of different levels.
And the quality evaluation value calculation unit 205 is configured to calculate the void influence values of the reverse threshold images at different levels, obtain a board quality evaluation value according to the void influence values and the tone consistency indexes, and evaluate the quality of the particle board by using the quality evaluation value.
Compared with the traditional technical scheme, the quality evaluation method disclosed by the invention has the advantages that the quality evaluation is carried out on the solid wood particle board based on the image processing, so that the detection efficiency and the detection precision can be improved; holes in the solid wood particle board are classified, the influence of the holes on the quality of the board is respectively considered, and a more accurate quality detection result can be effectively obtained.
The use of words such as "including," "comprising," "having," and the like in this disclosure is an open-ended term that means "including, but not limited to," and is used interchangeably therewith. The words "or" and "as used herein mean, and are used interchangeably with, the word" and/or, "unless the context clearly dictates otherwise. The word "such as" is used herein to mean, and is used interchangeably with, the phrase "such as but not limited to".
It should also be noted that the various components or steps may be broken down and/or re-combined in the methods and systems of the present invention. These decompositions and/or recombinations are to be considered equivalents of the present disclosure.
The above-mentioned embodiments are merely examples for clearly illustrating the present invention and do not limit the scope of the present invention. It will be apparent to those skilled in the art that other variations and modifications may be made in the foregoing description, and it is not necessary or necessary to exhaustively enumerate all embodiments herein. All designs identical or similar to the present invention are within the scope of the present invention.

Claims (4)

1. The method for evaluating the quality of the solid wood particle board based on image processing is characterized by comprising the following steps of:
carrying out semantic segmentation on the collected wood board section image to obtain a wood board section image;
converting the wood board section image to obtain a wood section HSV image, extracting a wood board tone channel image, calculating tone variance of the tone channel image, and taking the tone variance as a tone consistency index;
carrying out gray level conversion on the wood board section image to obtain a gray level image, and carrying out level division on the gray level image to obtain different level gray level images;
performing threshold segmentation on the gray level images of different levels to obtain threshold segmentation images, performing reverse threshold processing on the threshold segmentation images to obtain reverse threshold images of different levels, and taking the image with a pixel value of 1 in the reverse threshold images as a hole;
calculating the hole influence values of the reverse threshold images of different levels, obtaining a board quality evaluation value according to the hole influence values and the tone consistency indexes, and evaluating the quality of the particle board by using the quality evaluation value;
wherein, the calculation model of the wood board quality evaluation value is as follows:
Figure DEST_PATH_IMAGE002
in the formula
Figure DEST_PATH_IMAGE004
Of fingersIs the evaluation value of the tone consistency of the wood tone channel image,
Figure DEST_PATH_IMAGE006
is as follows
Figure DEST_PATH_IMAGE008
Influence values of holes in the image of the reverse threshold value of each level;
the holes comprise large holes and small holes; the calculation model of the influence value of the holes in the reverse threshold value image at each level is as follows:
Figure DEST_PATH_IMAGE010
in the formula
Figure DEST_PATH_IMAGE012
Is as follows
Figure 76397DEST_PATH_IMAGE008
The total number of small holes in the reverse threshold image of each level,
Figure DEST_PATH_IMAGE014
is as follows
Figure 649330DEST_PATH_IMAGE008
The total number of large holes in the individual level inverse threshold image,
Figure DEST_PATH_IMAGE016
first finger
Figure DEST_PATH_IMAGE018
The area of each large hole is larger than that of the corresponding hole,
Figure DEST_PATH_IMAGE020
first finger
Figure 868215DEST_PATH_IMAGE018
A large hole and the first
Figure DEST_PATH_IMAGE022
The shortest distance between the big holes,
Figure DEST_PATH_IMAGE024
the average area of all the large holes, the large hole is a connected domain with the area larger than the average area of all the connected domains, the small hole is a connected domain with the area not larger than the average area of all the connected domains,
Figure DEST_PATH_IMAGE026
wherein
Figure DEST_PATH_IMAGE028
Refers to the longest distance between large cavities.
2. The method for evaluating the quality of the solid wood particle board based on the image processing as claimed in claim 1, wherein the step of dividing the gray level image into layers comprises the steps of:
calculating a variance value of the gray scale image
Figure DEST_PATH_IMAGE030
Then, a straight line parallel to the horizontal direction is used to move downwards pixel by pixel from the upper part of the image, and after one pixel is moved each time, the intra-class variance at the two sides of the straight line is respectively calculated
Figure DEST_PATH_IMAGE032
And
Figure DEST_PATH_IMAGE034
；
Figure DEST_PATH_IMAGE036
，
Figure DEST_PATH_IMAGE038
and representing the inter-class variance, when the inter-class variance is maximum, minimizing the intra-class variance, taking the straight line as a dividing line, completing the hierarchical division of the upper image and the lower image, and dividing the lower image again until the gray image is divided, thereby finally obtaining the gray images with different levels.
3. The method for evaluating the quality of the solid wood particle board based on the image processing as claimed in claim 1, wherein the calculation model of the evaluation value of the color tone consistency is as follows:
Figure DEST_PATH_IMAGE040
in the formula
Figure DEST_PATH_IMAGE042
A reference numeral is given to each of the pixels,
Figure DEST_PATH_IMAGE044
represents the total number of pixels and the total number of pixels,
Figure DEST_PATH_IMAGE046
represents the first
Figure 264691DEST_PATH_IMAGE042
The value of each of the pixels is calculated,
Figure DEST_PATH_IMAGE048
representing the pixel mean.
4. The utility model provides a wood particle board material quality evaluation system based on image processing which characterized in that includes:
the wood slice image acquisition unit is used for performing semantic segmentation on the acquired wood board section image to obtain a wood board section image;
the tone consistency calculating unit is used for converting the wood board section image to obtain a wood section HSV image, extracting a wood board tone channel image, calculating tone variance of the tone channel image and taking the tone variance as a tone consistency index;
the gray image level division unit is used for carrying out gray level conversion on the wood board section image to obtain a gray image and carrying out level division on the gray image to obtain different level gray images;
the image reverse threshold processing unit is used for carrying out threshold segmentation on the gray level images of different levels to obtain threshold segmentation images, carrying out reverse threshold processing on the threshold segmentation images to obtain reverse threshold images of different levels, and taking the image with the pixel value of 1 in the reverse threshold images as a hole;
the quality evaluation value calculation unit is used for calculating the hole influence values of the reverse threshold images of different layers, obtaining a board quality evaluation value according to the hole influence values and the tone consistency indexes, and evaluating the quality of the particle board by using the quality evaluation value; wherein, the calculation model of the wood board quality evaluation value is as follows:
Figure DEST_PATH_IMAGE002A
in the formula
Figure 187385DEST_PATH_IMAGE004
Refers to the tone consistency evaluation value of the wood tone channel image,
Figure 947531DEST_PATH_IMAGE006
is as follows
Figure 201182DEST_PATH_IMAGE008
Influence values of holes in the image of the reverse threshold value of each level; the holes comprise large holes and small holes; the calculation model of the influence value of the holes in the reverse threshold value image at each level is as follows:
Figure DEST_PATH_IMAGE010A
in the formula
Figure 321454DEST_PATH_IMAGE012
Is as follows
Figure 177283DEST_PATH_IMAGE008
The total number of small holes in the reverse threshold image of each level,
Figure 108330DEST_PATH_IMAGE014
is as follows
Figure 312259DEST_PATH_IMAGE008
The total number of large holes in the level reverse threshold image is
Figure 986954DEST_PATH_IMAGE018
The area of each large hole is larger than that of the corresponding hole,
Figure 838236DEST_PATH_IMAGE020
first finger
Figure 720610DEST_PATH_IMAGE018
A large hole and the first
Figure 165498DEST_PATH_IMAGE022
The shortest distance between the big holes,
Figure 627572DEST_PATH_IMAGE024
the average area of all the large holes, the large hole is a connected domain with the area larger than the average area of all the connected domains, the small hole is a connected domain with the area not larger than the average area of all the connected domains,
Figure 208726DEST_PATH_IMAGE026
wherein
Figure 999352DEST_PATH_IMAGE028
Finger with large holeThe longest distance therebetween.
CN202111104111.3A 2021-09-22 2021-09-22 Image processing-based method and system for evaluating quality of solid wood particle board material Active CN113570599B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111104111.3A CN113570599B (en) 2021-09-22 2021-09-22 Image processing-based method and system for evaluating quality of solid wood particle board material

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111104111.3A CN113570599B (en) 2021-09-22 2021-09-22 Image processing-based method and system for evaluating quality of solid wood particle board material

Publications (2)

Publication Number Publication Date
CN113570599A true CN113570599A (en) 2021-10-29
CN113570599B CN113570599B (en) 2021-12-07

Family

ID=78173891

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111104111.3A Active CN113570599B (en) 2021-09-22 2021-09-22 Image processing-based method and system for evaluating quality of solid wood particle board material

Country Status (1)

Country Link
CN (1) CN113570599B (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115082478A (en) * 2022-08-23 2022-09-20 凤芯微电子科技(聊城)有限公司 Integrated circuit board quality sorting system
CN117934452A (en) * 2024-03-14 2024-04-26 中国建筑第五工程局有限公司 Rapid autoclaved concrete slab material quality detection method based on artificial intelligence

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101876641A (en) * 2009-04-30 2010-11-03 康宁股份有限公司 Be used for detecting the method and apparatus of the defective of glass plate
CN102886384A (en) * 2011-07-19 2013-01-23 宝山钢铁股份有限公司 Flatness defect identification method of 20-roller Sendzimir rolling mill based on support vector machine
CN205426713U (en) * 2015-10-15 2016-08-03 南京林业大学 Timber wood shavings and granule unit quality analysis appearance
CN106683093A (en) * 2017-01-12 2017-05-17 国家林业局北京林业机械研究所 Board appearance quality comprehensive quantitative evaluation method
CN106680299A (en) * 2017-01-12 2017-05-17 国家林业局北京林业机械研究所 Plate knurr appearance quality evaluation method
US20180238820A1 (en) * 2013-01-30 2018-08-23 Giatec Scientific Inc. Method and systems relating to construction material assessment

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101876641A (en) * 2009-04-30 2010-11-03 康宁股份有限公司 Be used for detecting the method and apparatus of the defective of glass plate
CN102886384A (en) * 2011-07-19 2013-01-23 宝山钢铁股份有限公司 Flatness defect identification method of 20-roller Sendzimir rolling mill based on support vector machine
US20180238820A1 (en) * 2013-01-30 2018-08-23 Giatec Scientific Inc. Method and systems relating to construction material assessment
CN205426713U (en) * 2015-10-15 2016-08-03 南京林业大学 Timber wood shavings and granule unit quality analysis appearance
CN106683093A (en) * 2017-01-12 2017-05-17 国家林业局北京林业机械研究所 Board appearance quality comprehensive quantitative evaluation method
CN106680299A (en) * 2017-01-12 2017-05-17 国家林业局北京林业机械研究所 Plate knurr appearance quality evaluation method

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115082478A (en) * 2022-08-23 2022-09-20 凤芯微电子科技(聊城)有限公司 Integrated circuit board quality sorting system
CN115082478B (en) * 2022-08-23 2022-11-18 凤芯微电子科技(聊城)有限公司 Integrated circuit board quality sorting system
CN117934452A (en) * 2024-03-14 2024-04-26 中国建筑第五工程局有限公司 Rapid autoclaved concrete slab material quality detection method based on artificial intelligence
CN117934452B (en) * 2024-03-14 2024-06-07 中国建筑第五工程局有限公司 Rapid autoclaved concrete slab material quality detection method based on artificial intelligence

Also Published As

Publication number Publication date
CN113570599B (en) 2021-12-07

Similar Documents

Publication Publication Date Title
CN107330889B (en) A kind of Chinese medicine tongue color coating colour automatic analysis method based on convolutional neural networks
CN113570599B (en) Image processing-based method and system for evaluating quality of solid wood particle board material
CN106056155B (en) Superpixel segmentation method based on boundary information fusion
CN103186904B (en) Picture contour extraction method and device
CN105894503B (en) A kind of restorative procedure of pair of Kinect plant colour and depth detection image
CN104574405B (en) A kind of coloured image threshold segmentation method based on Lab space
CN105893925A (en) Human hand detection method based on complexion and device
CN107146229B (en) Polyp of colon image partition method based on cellular Automation Model
CN111460894B (en) Intelligent car logo detection method based on convolutional neural network
CN113409267B (en) Pavement crack detection and segmentation method based on deep learning
CN103617596A (en) Image color style transformation method based on flow pattern transition
CN104268590A (en) Blind image quality evaluation method based on complementarity combination characteristics and multiphase regression
CN109492668A (en) MRI based on multichannel convolutive neural network not same period multi-mode image characterizing method
CN110222712B (en) Multi-special-item target detection algorithm based on deep learning
CN104658003A (en) Tongue image segmentation method and device
CN106529432A (en) Hand area segmentation method deeply integrating significance detection and prior knowledge
CN107392904A (en) A kind of partitioning algorithm of the medical image based on mathematical morphology
CN106355607A (en) Wide-baseline color image template matching method
CN104331877A (en) Color image edge detection method based on fusion color gradient
CN109376719B (en) Camera light response non-uniformity fingerprint extraction and comparison method based on combined feature representation
CN103136729B (en) Fuzzy vector morphological filtering method based on hypercomplex description
CN105513071B (en) A kind of topographic map symbols quality evaluating method
CN116402802A (en) Underwater image quality evaluation method based on color space multi-feature fusion
JP2013196681A (en) Method and device for extracting color feature
CN1852392A (en) Printing net-point-image dividing method based on moveable contour

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
GR01 Patent grant
GR01 Patent grant
TR01 Transfer of patent right
TR01 Transfer of patent right

Effective date of registration: 20240111

Address after: No. 20 Xinuo Road, Yudong Town, Haimen District, Nantong City, Jiangsu Province, 226100

Patentee after: Nantong Beilei exhibition cabinet manufacturing Co.,Ltd.

Address before: 226000 group 11, Xinyu village, Yudong Town, Haimen City, Nantong City, Jiangsu Province

Patentee before: Jiangsu Lvquan Decoration Engineering Co.,Ltd.