CN115165886A - Textile flame-retardant finishing quality detection method based on vision - Google Patents

Textile flame-retardant finishing quality detection method based on vision Download PDF

Info

Publication number
CN115165886A
CN115165886A CN202211075673.4A CN202211075673A CN115165886A CN 115165886 A CN115165886 A CN 115165886A CN 202211075673 A CN202211075673 A CN 202211075673A CN 115165886 A CN115165886 A CN 115165886A
Authority
CN
China
Prior art keywords
smoldering
time
region
area
transition
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
CN202211075673.4A
Other languages
Chinese (zh)
Other versions
CN115165886B (en
Inventor
潘忠国
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.)
Zibo Qihua Garment Co ltd
Original Assignee
Zibo Qihua Garment 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 Zibo Qihua Garment Co ltd filed Critical Zibo Qihua Garment Co ltd
Priority to CN202211075673.4A priority Critical patent/CN115165886B/en
Publication of CN115165886A publication Critical patent/CN115165886A/en
Application granted granted Critical
Publication of CN115165886B publication Critical patent/CN115165886B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/84—Systems specially adapted for particular applications
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N31/00—Investigating or analysing non-biological materials by the use of the chemical methods specified in the subgroup; Apparatus specially adapted for such methods
    • G01N31/12—Investigating or analysing non-biological materials by the use of the chemical methods specified in the subgroup; Apparatus specially adapted for such methods using combustion
    • 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
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/11—Region-based segmentation
    • 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
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/762—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using clustering, e.g. of similar faces in social networks
    • 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
    • G06T2207/10—Image acquisition modality
    • G06T2207/10024—Color image
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
    • Y02P90/30—Computing systems specially adapted for manufacturing

Abstract

The invention relates to the technical field of image processing, in particular to a textile flame-retardant finishing quality detection method based on vision. The method comprises the steps of determining the flame continuation time according to the flame continuation time on a fabric sample, shooting fabric sample images at set time intervals, determining the smoldering time according to the time for the smoldering area in the obtained gray level image to stop changing, determining the transition areas among a glowing combustion area, a pyrolysis carbonization area and a residual carbon area in the smoldering area according to the transition characteristics of each subdivided clustering area in the smoldering area, determining the duration time of the glowing combustion area, and finally finishing accurate judgment of the flame retardant quality of the fabric according to the obtained flame continuation time, the glowing time and the duration time of the glowing combustion area. The method has no waiting time when a gas analyzer is used for detecting the smoldering process, overcomes the defect that the traditional image processing method cannot accurately detect the smoldering process, and finally realizes the high-efficiency and accurate detection of the flame retardant quality of the fabric by a machine vision method.

Description

Textile flame-retardant finishing quality detection method based on vision
Technical Field
The invention relates to the technical field of image processing, in particular to a textile flame-retardant finishing quality detection method based on vision.
Background
The fabric is easy to cause fire in the processes of accumulation and use, so that the fabric needs to be subjected to flame retardant finishing, namely flame retardant treatment, so that the fabric has flame retardant capability, and the flame retardant quality is detected by a method of igniting the fabric after the flame retardant finishing.
In the flame retardant quality detection process, the fabric combustion comprises two processes of continuous combustion and smoldering, wherein the smoldering process is more difficult to detect due to the fact that naked flame does not exist, namely visible flame does not exist.
Although the prior art has a relevant means for identifying fire by using an image processing method, the specific process of smoldering cannot be accurately identified by using the current image identification method because flame does not exist in the smoldering process and the process is short, thereby finishing the judgment of the flame retardant quality of the fabric.
In conclusion, the prior art cannot realize the efficient and accurate detection of the flame retardant quality of the fabric.
Disclosure of Invention
The invention provides a method for detecting the flame-retardant finishing quality of a fabric based on vision, which is used for solving the problem that the flame-retardant quality of the fabric can not be efficiently and accurately detected in the prior art, and adopts the following technical scheme:
the invention relates to a textile flame-retardant finishing quality detection method based on vision, which comprises the following steps:
igniting the fabric sample, determining the flame duration on the fabric sample at the start of the ignition operation and determining the afterflame time of the fabric sample
Figure DEST_PATH_IMAGE001
；
Shooting a fabric sample at a set time interval from the end time of the afterflame time, synchronously acquiring an RGB (red, green and blue) image and an infrared image of the fabric sample, and carrying out graying processing on the acquired infrared image to obtain a grayscale image;
performing foreground segmentation on the shot RGB image to determine a fabric sample region, then performing foreground segmentation again on the fabric sample region to determine a region different from a normal fabric color on the fabric sample region, and determining a smoldering region on the corresponding gray level image according to the position of the determined region different from the normal fabric color, wherein the normal fabric color is the color of the fabric not burnt on the fabric sample;
calculating the difference value between the gray gradient values of different pixel points in the smoldering region, taking the absolute value of the difference value as the distance between different pixel points in the smoldering region, and clustering all the pixel points in the smoldering region to obtain a set number of clustering regions;
determining a macro transition index of each clustering region and a micro transition index of each pixel point in the clustering region, and determining a region transition index of the clustering region according to the macro transition index of the clustering region and the micro transition index of each pixel point in the clustering region;
the microscopic transition index at the pixel point is:
Figure DEST_PATH_IMAGE003
wherein the content of the first and second substances,
Figure 520879DEST_PATH_IMAGE004
is the microscopic transition index at the ith pixel point in the smoldering region,
Figure DEST_PATH_IMAGE005
the size of a set neighborhood taking the ith pixel point as a central point, N is the side length of the set neighborhood,
Figure 168898DEST_PATH_IMAGE006
to set the gray gradient value of the jth pixel point in the neighborhood with the ith pixel point as the center point,
Figure DEST_PATH_IMAGE007
the average value of the gray gradient values of all pixel points in the set neighborhood by taking the ith pixel point as a central point,
Figure 947146DEST_PATH_IMAGE008
the standard deviation of the gray gradient values of all pixel points in the set neighborhood by taking the ith pixel point as a central point,
Figure DEST_PATH_IMAGE009
and
Figure 313405DEST_PATH_IMAGE010
respectively arranging the upper quartile and the lower quartile in sequence from small to big according to the gray gradient values of all pixel points in a set neighborhood taking the ith pixel point as a central point;
the macroscopic transition index of the clustering area is as follows:
Figure 916425DEST_PATH_IMAGE012
wherein, the first and the second end of the pipe are connected with each other,
Figure DEST_PATH_IMAGE013
is the macroscopic transition index of the z-th clustering interval,
Figure 684529DEST_PATH_IMAGE014
and
Figure DEST_PATH_IMAGE015
respectively the maximum value and the minimum value of the gray gradient value in the clustering interval,
Figure 878750DEST_PATH_IMAGE016
the number of the pixel points in the clustering interval,
Figure DEST_PATH_IMAGE017
、
Figure 735236DEST_PATH_IMAGE018
and
Figure DEST_PATH_IMAGE019
respectively are distance values between the centroid of the clustering interval and the centroids of three other clustering intervals which are closest to the centroid;
the region transition index of the clustering region is as follows:
Figure DEST_PATH_IMAGE021
wherein the content of the first and second substances,
Figure 1001DEST_PATH_IMAGE022
a region transition index representing the z-th clustering region,
Figure DEST_PATH_IMAGE023
represents the sum of the microscopic transition indexes at all pixel points in the z-th clustering region,
Figure 889191DEST_PATH_IMAGE024
representing the average value of the microscopic transition indexes at all the pixel points in the z-th clustering area;
calculating the mean of the regional transition indexes of all clustered regions
Figure DEST_PATH_IMAGE025
Then, the regional transition index is larger than the mean value
Figure 844859DEST_PATH_IMAGE025
The clustering area is marked as a transition area to be judged;
clustering all transition areas to be judged to obtain a plurality of cluster clusters, if all the transition areas to be judged have adjacent transition areas to be judged and the number of the cluster clusters is 2, a glowing combustion area between a pyrolysis carbonization area and a residual carbon area exists in a smoldering area, and the duration time of the glowing combustion area from the end time of the continuous combustion time is recorded
Figure 123394DEST_PATH_IMAGE026
；
Detecting the number of pixel points in the smoldering region, and recording the time from the end moment of the continuous combustion time to the moment when the number of the pixel points in the smoldering region does not change any more as the smoldering time
Figure DEST_PATH_IMAGE027
；
Judging the time of continuous combustion
Figure 396112DEST_PATH_IMAGE001
And smoldering time
Figure 545334DEST_PATH_IMAGE027
Whether the time exceeds the corresponding national standard specified time, if so, the time of continuous combustion
Figure 550199DEST_PATH_IMAGE001
And smoldering time
Figure 316029DEST_PATH_IMAGE027
If at least one of the flame-retardant fabrics exceeds the corresponding national standard specified time, the flame-retardant quality of the fabric is unqualified, otherwise, the duration of the scorching combustion zone is continued
Figure 801893DEST_PATH_IMAGE026
And smoldering time
Figure 336780DEST_PATH_IMAGE027
Judging the ratio of (A) to (B):
duration of glowing combustion zone
Figure 981387DEST_PATH_IMAGE026
And smoldering time
Figure 234514DEST_PATH_IMAGE027
If the ratio of (A) to (B) is less than the first ratio, the flame-retardant quality of the fabric is perfect, if the glowing combustion zone duration is long
Figure 521139DEST_PATH_IMAGE026
And smoldering time
Figure 910532DEST_PATH_IMAGE027
Is not less than the first ratio and is less than the second ratio, the flame retardant quality of the fabric is excellent if the glowing combustion zone duration is
Figure 257200DEST_PATH_IMAGE026
And smoldering time
Figure 732044DEST_PATH_IMAGE027
The ratio of (a) to (b) is not less than the second ratio, the flame retardant quality of the fabric is good;
the first ratio is less than the second ratio.
Further, the method for judging whether the transition area to be judged has an adjacent transition area to be judged comprises the following steps:
if two pixel points are respectively positioned in eight adjacent areas of the opposite side, the two pixel points are adjacent; and if the pixel point in the transition region to be judged is adjacent to any pixel point in other transition regions to be judged, the two transition regions to be judged are adjacent.
Further, the specific method for determining the duration of the scorching combustion zone is as follows:
shooting the fabric sample at this time, analyzing and determining that no burning hot combustion area exists according to the shot image, analyzing and determining that the burning hot combustion area exists according to the shot image after each shooting, and recording the shooting times of the shooting at this time
Figure 288271DEST_PATH_IMAGE028
Then calculating the duration of the burning zone
Figure DEST_PATH_IMAGE029
Wherein T is the set time interval.
Further, the specific method for determining the smoldering time comprises the following steps:
when the fabric sample is shot at this time and the number of pixel points contained in the smoldering area determined according to the shot image analysis is the same as the number of pixel points contained in the smoldering area determined corresponding to the previous shooting, recording the shooting of the shooting at this timeNumber of shots taken
Figure 860067DEST_PATH_IMAGE030
And continuing to shoot the fabric sample and analyzing and determining the number of pixel points contained in the smoldering area according to the shot image, and calculating the smoldering time if the number of the pixel points contained in the smoldering area is not changed under the condition of continuously setting the shooting times
Figure DEST_PATH_IMAGE031
Wherein T is the set time interval.
Further, the first ratio is
Figure 439953DEST_PATH_IMAGE032
The second ratio is
Figure DEST_PATH_IMAGE033
。
The beneficial effects of the invention are as follows:
the method comprises the steps of shooting a fabric sample, synchronously obtaining an RGB (red, green and blue) image of the fabric sample and a gray image after graying processing of an infrared image, determining a smoldering region in the gray image by means of the RGB image, comprehensively determining the interval transition index of each clustering interval according to the microscopic transition index of each pixel point in the smoldering region and the macroscopic transition index of each clustering interval subdivided in the smoldering region, determining whether a glowing combustion area between a pyrolysis carbonization area and a residual carbon area exists in the smoldering region according to the position distribution of the transition area to be determined, correspondingly obtaining the duration of the glowing combustion area, determining the smoldering time according to whether the size of the smoldering region stops changing, determining the continuous combustion time according to the flame duration after the fabric sample is ignited, finishing flame retardant quality determination according to the determined continuous combustion time, smoldering time and glowing combustion area duration, and avoiding the problem that a gas analyzer in the prior art cannot accurately detect the smoldering process, and finally realizing the accurate flame retardant quality detection of the fabric by using the conventional image processing method.
Drawings
FIG. 1 is a flow chart of the vision-based method for detecting the flame-retardant finishing quality of the fabric according to the invention;
FIG. 2 is a schematic illustration of a pyrolysis carbonization zone, a glowing combustion zone, and a residual carbon zone in a smoldering zone of the present invention;
FIG. 3 is a schematic representation of the transition region between the hot burn zone and the pyrolytic carbonization zone and the transition region between the hot burn zone and the residual carbon zone in the smoldering zone of the present invention.
Detailed Description
The conception of the invention is as follows:
the method comprises the steps of firstly detecting the continuous combustion time of the cloth through a photosensitive device, starting the detection of the smoldering process at the continuous combustion ending moment, determining a smoldering area through a foreground segmentation method in the smoldering detection process, determining the whole smoldering time by taking the time point of stopping the expansion of the smoldering area as the smoldering ending time, determining whether a transition area between a glowing combustion area and a residual carbon area and a transition area between the glowing combustion area and a pyrolysis carbonization area exist in the smoldering area simultaneously through the microscopic transition index of a single pixel point and the macroscopic transition index of each subdivided area in the smoldering detection process, determining the existence time of the glowing combustion area, and finally determining the good and bad specific flame retardant quality of the cloth through comparing the proportional relation between the whole smoldering time and the existence time of the glowing combustion area on the premise that the continuous combustion time and the whole smoldering time meet requirements.
The method for detecting the flame-retardant finishing quality of the textile based on vision is described in detail below with reference to the accompanying drawings and examples.
The method comprises the following steps:
the embodiment of the method for detecting the flame-retardant finishing quality of the fabric based on vision has the overall flow as shown in figure 1, and comprises the following specific processes:
igniting the fabric sample, determining the afterflame time of the fabric sample, starting to acquire an infrared image and an RGB (red, green and blue) image of the fabric sample at the moment of finishing the afterflame, and determining a smoldering area.
The method comprises the steps of cutting out a fabric sample with a proper size at the edge part of a fabric to be detected according to the mode that the side length of the sample is parallel to the warp direction or the weft direction of the fabric, fixing one end of the sample on a flame retardant test support to enable the fabric sample to naturally fall, igniting the fabric sample from the bottom end of the fabric sample, wherein the ignition time is determined as set ignition time, the specific value of the ignition time is determined according to the specific material and density of the fabric, and the set ignition time is preferably 2s in the embodiment.
Because the fabric is subjected to the flame-retardant finishing treatment, the fabric can delay the ignition process after contacting with flame and gradually stop burning after being far away from the flame. The combustion stopping generally needs two processes of afterflame and smoldering, the afterflame is flaming combustion, so the time is easy to determine, the time can be determined by measuring the flame duration through a photosensitive device after the set ignition time of 2s is finished, and the afterflame time is recorded as
Figure 464409DEST_PATH_IMAGE001
Smoldering is a slow burn without visible light and is therefore difficult to identify and detect.
Based on the arranged photosensitive device such as a photoelectric sensor, when flaming combustion is stopped after flame combustion is finished, the photosensitive device does not receive light emitted by flame any more, and at the moment, the thermal infrared imager is automatically triggered to shoot infrared images of the sample at a set time interval T, in the embodiment, the set time interval is preferably 0.1s, that is, one infrared image is shot every 0.1 s. Meanwhile, the RGB images of the sample are synchronously acquired at the same position also at set time intervals using an industrial camera. When an image is obtained, the burning position needs to be arranged at a more central position of the image for facilitating subsequent analysis.
Because the stable smoldering combustion and the flaming combustion can be simultaneously maintained and the afterflame is also continued generally only under the flaming condition of high concentration and high oxygen supply, and the sample is far away from the flame after ignition, the afterflame and the smoldering combustion can not occur simultaneously, the image is shot from the end of the afterflame to analyze and time the smoldering combustion, and the accuracy of measuring the smoldering combustion can not be influenced.
In order to ensure the accuracy of subsequent analysis, gaussian filtering is used for carrying out convolution with each channel of the RGB image of the sample, and the interference of noise is eliminated. The infrared image is a pseudo-color image, so that the infrared image is converted into a gray image, and the gray value of the pixel point corresponding to the position with higher temperature is larger.
Because in smoldering process, the pyrolysis carbonization zone can produce smog, and the region that has smog in the image can be comparatively fuzzy, for avoiding smog to the influence of follow-up analysis, uses dark channel to test algorithm earlier and removes smog processing to the RGB image of sample.
Of course, in other embodiments, only one of the gaussian filtering and the smoke removal processing may be performed or neither may be performed according to the requirement of the detection accuracy.
After the RGB image of the sample is acquired, the captured sample RGB image is divided into two parts, corresponding to the background region and the sample region, respectively, using OTSU law enforcement. Since the position where the fabric is burned by the flame and the position where smoldering occurs are generally black and have a significant difference from the color of the unburned fabric region, the OTSU law method is used again for the determined sample region to mark out the darker part, namely the region burned by the flame and the region where smoldering reaction occurs.
Based on the areas after being burned by flame and the areas in which smoldering reactions are occurring, which are determined in the sample RGB image, the smoldering areas can be determined in the gray scale image obtained by graying the sample infrared image.
And secondly, clustering the pixels in the smoldering region according to the gray gradient values of all the pixels in the smoldering region, subdividing the smoldering region into a set number of clustering regions, reclassifying the clustering regions according to the transition characteristics of all the clustering regions, and determining whether a transition region between the scorching combustion region and the pyrolysis carbonization region and a transition region between the scorching combustion region and the residual carbon region exist simultaneously, so as to determine whether the scorching combustion region exists and the duration time of the scorching combustion region, and determine the duration time of the whole smoldering process.
As shown in fig. 2, the smoldering zone, although not having a definite regular shape, may be divided into three parts, from bottom to top, which are a residual carbon zone, a scorching combustion zone, and a pyrolysis carbonization zone, respectively.
The residual carbon area corresponds to the residue part after combustion, and is at a lower temperature and in a continuous cooling state; the glowing combustion area is a part of solid carbon for gas-solid two-phase combustion, the reaction is most violent, and the temperature is highest; the pyrolysis carbonization area is the part of the material which is thermally decomposed, and the temperature is higher. Although the three parts do not have exact temperature as a division threshold value, the temperature in each part is uniform, the temperature difference between the scorching combustion area and the other two parts is large, and transition areas with gradually changing temperatures exist between the scorching combustion area and the other two parts as shown in fig. 3, so that the transition areas can be identified according to the transition characteristics, and whether the scorching combustion area exists or not and the duration time are judged according to the transition areas. The method comprises the following specific steps:
1. a microscopic transition index at each pixel point in the smoldering region is determined.
Because the transition region between the pyrolysis carbonization region and the scorching combustion region and the transition region between the scorching combustion region and the residual carbon region have obvious temperature changes, the gray value of the transition region in the gray image has obvious gray change.
Firstly, determining the gray gradient value of each pixel point in the smoldering area, wherein the gray gradient value can be determined by adopting any available method, preferably, in the embodiment, a sobel operator is convoluted with the gray image to obtain a gray gradient image, and then, the gray gradient value of each pixel point in the smoldering area is obtained.
Then, each pixel point in the smoldering region is used as a center to determine a set neighborhood, the set neighborhood is a square region with the side length of N, and the mean value of the gray gradient values of all the pixel points in the neighborhood is recorded as
Figure 92837DEST_PATH_IMAGE034
Standard deviation is recorded as
Figure 194173DEST_PATH_IMAGE008
And sequencing the gray gradient values of the pixel points in the set neighborhood from small to large to determine the upper quartile
Figure 882643DEST_PATH_IMAGE009
And lower quartile
Figure 66500DEST_PATH_IMAGE010
And thus determining the microscopic transition index of each pixel:
Figure 233039DEST_PATH_IMAGE003
wherein the content of the first and second substances,
Figure 451530DEST_PATH_IMAGE004
is the microscopic transition index at the ith pixel point in the smoldering region,
Figure 983006DEST_PATH_IMAGE005
the size of a set neighborhood taking the ith pixel point as a central point, N is the side length of the set neighborhood,
Figure 654159DEST_PATH_IMAGE006
setting the gray gradient value of the jth pixel point in the neighborhood by taking the ith pixel point as a central point,
Figure 889968DEST_PATH_IMAGE007
the average value of the gray gradient values of all pixel points in the set neighborhood taking the ith pixel point as a central point,
Figure 962966DEST_PATH_IMAGE008
in order to set the standard deviation of the gray scale gradient values of all pixel points in the neighborhood by taking the ith pixel point as the center point,
Figure 724730DEST_PATH_IMAGE009
and
Figure 883179DEST_PATH_IMAGE010
the gray gradient values of all pixel points in the set neighborhood taking the ith pixel point as a central point are respectively an upper quartile and a lower quartile which are sequenced from small to big.
Microscopic transition index
Figure 391521DEST_PATH_IMAGE004
In the step (1), the first step,
Figure DEST_PATH_IMAGE035
and
Figure 912501DEST_PATH_IMAGE008
all represent the fluctuation degree of the gray gradient values of all pixel points in the set neighborhood with the ith pixel point as the center point,
Figure 113675DEST_PATH_IMAGE036
the difference degree of the gray gradient values of all pixel points in the set neighborhood taking the ith pixel point as the center point is expressed when
Figure 759420DEST_PATH_IMAGE036
、
Figure 71453DEST_PATH_IMAGE035
And
Figure 121973DEST_PATH_IMAGE008
the larger the value of (a), the larger the difference degree and the larger the fluctuation degree of the gray scale gradient values of all the pixel points in the set neighborhood taking the ith pixel point as the center point are, and the corresponding microscopic transition index
Figure 494048DEST_PATH_IMAGE004
The larger the size, the more likely the pixel is in the transition region.
2. And clustering the pixels in the smoldering area according to the gray gradient value of each pixel point in the smoldering area, subdividing the smoldering area into a set number of clustering areas, and determining the macroscopic transition index of each clustering area.
The absolute value of the difference of the gray gradient values of different pixel points in the smoldering region is used as the distance between different pixel points, all the pixel points in the smoldering region are clustered to obtain a set number of clustering regions, in this embodiment, it is preferable to complete clustering by using a K-means clustering algorithm, and the set number is 64, in other embodiments, other existing arbitrary clustering algorithms can be adopted, and other values can be taken for the set number.
Because the clustering is to divide the pixels with similar gray gradient values into one region, the gray values in three parts of a pyrolysis carbonization region, a scorching-heat combustion region and a residual carbon region in the smoldering region are uniform, and only the region where the scorching-heat combustion region transits to the other two regions has very obvious gray change, after clustering, the clustering regions corresponding to the three parts of the pyrolysis carbonization region, the scorching-heat combustion region and the residual carbon region in the smoldering region contain more pixels and are far away from the adjacent clustering regions. Then each region can be analyzed individually based on this.
Recording the maximum value of the gray gradient value in a clustering region as
Figure 627089DEST_PATH_IMAGE014
The minimum value of the gray gradient value is recorded as
Figure 742813DEST_PATH_IMAGE015
The number of pixels is recorded as
Figure 379331DEST_PATH_IMAGE016
. Determining the mass center of each clustering region, calculating Euclidean distances between the mass center of the clustering region and the mass centers of other clustering regions, and respectively recording the three nearest Euclidean distances as
Figure DEST_PATH_IMAGE037
. Establishing the macro transition index of the clustering interval based on the related indexes
Figure 984624DEST_PATH_IMAGE013
：
Figure 339382DEST_PATH_IMAGE012
Wherein the content of the first and second substances,
Figure 544884DEST_PATH_IMAGE013
is the macroscopic transition index of the z-th clustering interval,
Figure 35908DEST_PATH_IMAGE014
and
Figure 749786DEST_PATH_IMAGE015
respectively the maximum value and the minimum value of the gray gradient value in the clustering interval,
Figure 857419DEST_PATH_IMAGE016
the number of the pixel points in the clustering interval,
Figure 314945DEST_PATH_IMAGE017
、
Figure 926055DEST_PATH_IMAGE018
and
Figure 873152DEST_PATH_IMAGE019
respectively are the distance values between the centroid of the clustering interval and the centroids of three other clustering intervals which are closest to each other.
Macroscopic transition index
Figure 202502DEST_PATH_IMAGE013
In the step (1), the first step,
Figure 201069DEST_PATH_IMAGE038
the difference degree of the gray gradient values in the clustering region is represented, when the number of the pixel points in the clustering region is smaller, the difference of the gray gradient values corresponding to the pixel points is larger, and the Euclidean distance between the clustering region and other clustering regions is closer, the clustering region is more likely to be alignedA transition region with a gradual temperature change.
3. And determining the region transition index of each clustering region.
Due to microscopic transition index
Figure 932265DEST_PATH_IMAGE004
From the microscopic angle, the transition condition of the gray gradient value distribution near each pixel point is evaluated, and the macroscopic transition index
Figure 253525DEST_PATH_IMAGE013
From the macroscopic view, the transition condition of the gray gradient value numerical value in each similar clustering region is evaluated, so that the two angles are combined to construct the clustering region transition index
Figure 804592DEST_PATH_IMAGE022
Comprehensively evaluating the transition situation in each area:
Figure 869500DEST_PATH_IMAGE021
wherein, the first and the second end of the pipe are connected with each other,
Figure 455202DEST_PATH_IMAGE022
a region transition index representing the z-th clustering region,
Figure 885046DEST_PATH_IMAGE013
represents the macroscopic transition index of the z-th clustering region,
Figure 985726DEST_PATH_IMAGE016
the number of pixel points in the z-th clustering area is represented,
Figure 789079DEST_PATH_IMAGE023
represents the sum of the microscopic transition indexes at all pixel points in the z-th clustering region,
Figure 229287DEST_PATH_IMAGE024
denotes the z thAverage of the microscopic transition indices at all pixel points within the clustered region.
When the microscopic transition indexes of all pixel points in the clustering region
Figure 689088DEST_PATH_IMAGE004
The larger the cluster area is, the total number of pixel points in the cluster area is
Figure 73801DEST_PATH_IMAGE016
The smaller, and the macroscopic transition index of the cluster region
Figure 746091DEST_PATH_IMAGE013
When the larger the cluster area is, the area transition index of the cluster area is
Figure 306385DEST_PATH_IMAGE022
The larger. And each clustering area in the smoldering area corresponds to an obtained area transition index.
4. Determining a transition region according to the regional transition index of each clustering region, thereby determining whether a burning region exists and the duration of the burning region, and simultaneously determining the duration of the whole smoldering process.
The mean value of the regional transition indexes of all the clustering regions is recorded as
Figure 877700DEST_PATH_IMAGE025
Then, the regional transition index is larger than the mean value
Figure 156234DEST_PATH_IMAGE025
The cluster area is marked as a transition area to be judged.
If two pixel points are respectively located in eight adjacent areas of the opposite side, the two pixel points are called to be adjacent. And if the pixel point in the obtained transition region to be judged is adjacent to any pixel point in other transition regions to be judged, the two transition regions to be judged are called to be adjacent.
Taking the centroids of all the transition areas to be determined of the adjacent transition areas to be determined, using DBSCAN clustering algorithm with 3 as the minimum number of points, and taking each centroid and other centroidsThe median of the Euclidean distances between them is the radius of the neighborhood, clustering is carried out, and the number of clusters obtained by clustering is recorded as
Figure DEST_PATH_IMAGE039
。
Since the residual carbon zone, the scorching combustion zone and the pyrolysis carbonization zone in the smoldering zone are divided by two transition zones, the following rules are provided:
if each transition area to be judged in the smoldering area has an adjacent transition area to be judged, and the number of clusters obtained by clustering
Figure 366636DEST_PATH_IMAGE039
When the number is 2, the corresponding positions of the transition areas to be determined are transition areas of the burning combustion area to the other two areas, and the transition areas to be determined are distributed in two bands far away from each other, as shown in fig. 3. In the smoldering zone, the area between the two transition areas is a burning combustion area, and the output result is that the burning combustion area exists.
If the two conditions are not met, the transition region from the burning combustion region to the other two regions does not exist, accordingly, the burning combustion region does not exist, and the output result is that the burning combustion region does not exist.
Whether a glowing combustion zone exists or not can be judged to be smoldering, and when the glowing combustion zone exists, smoldering reaction is still violent; and when the glowing combustion zone does not exist, the smoldering reaction is about to end, namely the overall smoldering time of the fabric sample is about to reach the timing stopping time.
When the fabric sample is shot and analyzed at this time, the result is determined to be that no burning hot combustion area exists, and the analysis results are all that the burning hot combustion area exists after each shooting, the shooting times of the shooting are recorded
Figure 781437DEST_PATH_IMAGE028
And recording the number of pixel points in the smoldering area at the moment.
Continuously shooting the fabric sample according to a set time interval and correspondingly determining a smoldering area after each shootingThe number of the pixels in the smoldering area is equal to the number of the pixels in the smoldering area corresponding to the last shooting, and the shooting times are recorded
Figure 786302DEST_PATH_IMAGE030
. And continuing to shoot and judge the number of the pixel points contained in the smoldering region in the image, and if the result corresponding to the continuously set shooting times is that the number of the pixel points contained in the smoldering region is not changed any more or the variation is smaller than the threshold value of the change of the number of the pixel points, judging that the smoldering is finished. At this time, the duration of the smoldering process is output, namely the smoldering time
Figure 20974DEST_PATH_IMAGE031
Duration of glowing combustion zone in smoldering combustion
Figure 769487DEST_PATH_IMAGE029
. In this embodiment, it is preferable to continuously set the number of times of shooting to 10 and the threshold value for changing the number of pixels to 5, and in other embodiments, the two values may be set to other values according to the requirement of detection accuracy.
And step three, finishing the judgment of the flame retardant quality of the fabric according to the afterflame time and smoldering time of the fabric sample and the duration of a glowing combustion area in the smoldering process.
Firstly, the time of continuous combustion is judged
Figure 35865DEST_PATH_IMAGE001
And smoldering time
Figure 477210DEST_PATH_IMAGE027
Whether the flame retardant quality of the fabric is not greater than the time specified by the national standard, if at least one of the flame retardant quality and the time specified by the corresponding national standard is greater than the time specified by the national standard, the flame retardant quality of the fabric is unqualified.
If intermittent combustion time
Figure 464758DEST_PATH_IMAGE001
And smoldering time
Figure 548120DEST_PATH_IMAGE027
Is not longer than the time specified by the corresponding national standard, the duration of the burning zone is continued
Figure 937514DEST_PATH_IMAGE026
And smoldering time
Figure 284181DEST_PATH_IMAGE027
Is judged if the duration of the burning zone is hot
Figure 759025DEST_PATH_IMAGE026
And smoldering time
Figure 852270DEST_PATH_IMAGE027
If the ratio of (A) to (B) is less than the first ratio, the flame-retardant quality of the fabric is perfect, if the glowing combustion zone duration is long
Figure 96170DEST_PATH_IMAGE026
And smoldering time
Figure 613739DEST_PATH_IMAGE027
Is not less than the first ratio and is less than the second ratio, the flame retardant quality of the fabric is excellent if the glowing combustion zone duration is
Figure 575879DEST_PATH_IMAGE026
And smoldering time
Figure 204306DEST_PATH_IMAGE027
Is not less than the second ratio, the flame retardant quality of the fabric is good.
The first ratio is smaller than the second ratio, the specific value of the first ratio and the second ratio is determined according to the specific requirement on the flame retardant quality of the fabric, and the first ratio is preferably selected in this embodiment
Figure 302712DEST_PATH_IMAGE032
The second ratio is
Figure 991182DEST_PATH_IMAGE033
。
Thus, the determination of the flame retardant quality of the fabric is completed.
The above-mentioned embodiments are only used to illustrate the technical solutions of the present application, and not to limit the same; although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; such modifications or substitutions do not cause the essential features of the corresponding technical solutions to depart from the scope of the technical solutions of the embodiments of the present application, and are intended to be included within the scope of the present application.

Claims (5)

1. A textile flame-retardant finishing quality detection method based on vision is characterized by comprising the following steps:
igniting the fabric sample, determining the flame duration on the fabric sample at the start of the ignition operation and determining the afterflame time of the fabric sample
Figure 176199DEST_PATH_IMAGE001
；
Shooting a fabric sample at a set time interval from the end time of the afterflame time, synchronously acquiring an RGB (red, green and blue) image and an infrared image of the fabric sample, and carrying out graying processing on the acquired infrared image to obtain a grayscale image;
performing foreground segmentation on the shot RGB image to determine a fabric sample region, then performing foreground segmentation on the fabric sample region again to determine a region different from a normal fabric color on the fabric sample region, and determining a smoldering region on a corresponding gray image according to the position of the determined region different from the normal fabric color, wherein the normal fabric color is the color of an unburnt fabric on the fabric sample;
calculating the difference value between the gray gradient values of different pixel points in the smoldering region, taking the absolute value of the difference value as the distance between different pixel points in the smoldering region, and clustering all the pixel points in the smoldering region to obtain a set number of clustering regions;
determining a macroscopic transition index of each clustering region and a microscopic transition index of each pixel point in the clustering region, and determining a region transition index of the clustering region according to the macroscopic transition index of the clustering region and the microscopic transition index of each pixel point in the clustering region;
the microscopic transition index at the pixel point is:
Figure 608317DEST_PATH_IMAGE002
wherein the content of the first and second substances,
Figure 847317DEST_PATH_IMAGE003
is the microscopic transition index at the ith pixel point in the smoldering region,
Figure 706688DEST_PATH_IMAGE004
the size of a set neighborhood taking the ith pixel point as a central point, N is the side length of the set neighborhood,
Figure 377841DEST_PATH_IMAGE005
to set the gray gradient value of the jth pixel point in the neighborhood with the ith pixel point as the center point,
Figure 348071DEST_PATH_IMAGE006
the average value of the gray gradient values of all pixel points in the set neighborhood by taking the ith pixel point as a central point,
Figure 421069DEST_PATH_IMAGE007
the standard deviation of the gray gradient values of all pixel points in the set neighborhood by taking the ith pixel point as a central point,
Figure 451342DEST_PATH_IMAGE008
and
Figure 609791DEST_PATH_IMAGE009
respectively an upper quartile and a lower quartile which are obtained by sequencing the gray gradient values of all pixel points in a set neighborhood with the ith pixel point as a central point from small to large;
the macro transition index of the clustering region is as follows:
Figure 383712DEST_PATH_IMAGE010
wherein the content of the first and second substances,
Figure 579726DEST_PATH_IMAGE011
is the macroscopic transition index of the z-th clustering interval,
Figure 780900DEST_PATH_IMAGE012
and
Figure 161066DEST_PATH_IMAGE013
respectively the maximum value and the minimum value of the gray gradient value in the clustering interval,
Figure 473098DEST_PATH_IMAGE014
the number of the pixels in the clustering interval is,
Figure 520689DEST_PATH_IMAGE015
、
Figure 892764DEST_PATH_IMAGE016
and
Figure 619281DEST_PATH_IMAGE017
respectively are distance values between the centroid of the clustering interval and the centroids of three other clustering intervals which are closest to the centroid;
the region transition index of the clustering region is as follows:
Figure 732074DEST_PATH_IMAGE018
wherein, the first and the second end of the pipe are connected with each other,
Figure 368592DEST_PATH_IMAGE019
a region transition index representing the z-th clustering region,
Figure 645990DEST_PATH_IMAGE020
represents the sum of the microscopic transition indexes at all pixel points in the z-th clustering region,
Figure 748DEST_PATH_IMAGE021
representing the average value of the microscopic transition indexes at all the pixel points in the z-th clustering area;
calculating the mean of the regional transition indexes of all clustered regions
Figure 513638DEST_PATH_IMAGE022
Then, the regional transition index is larger than the mean value
Figure 535820DEST_PATH_IMAGE022
The clustering area is marked as a transition area to be judged;
clustering all transition areas to be judged to obtain a plurality of cluster clusters, if all the transition areas to be judged have adjacent transition areas to be judged and the number of the cluster clusters is 2, then a glowing combustion area between a pyrolysis carbonization area and a residual carbon area exists in a smoldering area, and the duration time of the glowing combustion area from the end time of the continuous combustion time is recorded
Figure 390644DEST_PATH_IMAGE023
；
Detecting the number of the pixels in the smoldering region, and recording the time from the end moment of the continuous combustion time to the moment when the number of the pixels in the smoldering region does not change any more as the smoldering time
Figure 94682DEST_PATH_IMAGE024
；
Judging the time of continuous combustion
Figure 286629DEST_PATH_IMAGE001
And smoldering time
Figure 163318DEST_PATH_IMAGE024
Whether the time exceeds the corresponding national standard specified time, if so, the time of continuous combustion
Figure 48097DEST_PATH_IMAGE001
And smoldering time
Figure 377448DEST_PATH_IMAGE024
If at least one of the flame-retardant materials exceeds the corresponding national standard specified time, the flame-retardant quality of the fabric is unqualified, otherwise, the duration of the scorching burning zone is continued
Figure 638665DEST_PATH_IMAGE023
And smoldering time
Figure 369860DEST_PATH_IMAGE024
Judging the ratio of (A) to (B):
duration of glowing combustion zone
Figure 691120DEST_PATH_IMAGE023
And smoldering time
Figure 239258DEST_PATH_IMAGE024
If the ratio of (A) to (B) is less than the first ratio, the flame-retardant quality of the fabric is perfect, if the glowing combustion zone duration is long
Figure 569745DEST_PATH_IMAGE023
And smoldering time
Figure 155447DEST_PATH_IMAGE024
Is not less than the firstThe ratio is less than the second ratio, the flame retardant quality of the fabric is excellent if the duration of the scorching burn zone is long
Figure 382029DEST_PATH_IMAGE023
And smoldering time
Figure 685971DEST_PATH_IMAGE024
The ratio of (a) to (b) is not less than the second ratio, the flame retardant quality of the fabric is good;
the first ratio is less than the second ratio.
2. The visual-based fabric flame-retardant finishing quality detection method as claimed in claim 1, wherein the method for judging whether the transition area to be judged has an adjacent transition area to be judged is as follows:
if two pixel points are respectively positioned in eight adjacent areas of the opposite side, the two pixel points are adjacent; and if the pixel point in the transition region to be judged is adjacent to any pixel point in other transition regions to be judged, the two transition regions to be judged are adjacent.
3. The vision-based method for detecting the flame-retardant finishing quality of the textile fabric according to claim 1 or 2, wherein the specific method for determining the duration of the burning zone is as follows:
shooting the fabric sample at this time, analyzing and determining that no burning area exists according to the shot image, analyzing and determining that the burning area exists according to the shot image after each shooting, and recording the shooting times of the shooting
Figure 288991DEST_PATH_IMAGE025
Then calculating the duration of the burning zone
Figure 463620DEST_PATH_IMAGE026
Wherein T is the set time interval.
4. The vision-based method for detecting the flame-retardant finishing quality of the textile according to claim 1 or 2, characterized in that the specific method for determining the smoldering time is as follows:
when the fabric sample is shot at this time and the number of pixel points contained in the smoldering area determined according to the shot image analysis is the same as the number of pixel points contained in the smoldering area determined corresponding to the previous shooting, recording the shooting times of the shooting at this time
Figure 126683DEST_PATH_IMAGE027
And continuing to shoot the fabric sample and analyzing and determining the number of pixel points contained in the smoldering area according to the shot image, and calculating the smoldering time if the number of the pixel points contained in the smoldering area is not changed under the condition of continuously setting the shooting times
Figure 389693DEST_PATH_IMAGE028
Wherein T is the set time interval.
5. The vision-based method for detecting the flame-retardant finishing quality of textile fabrics of claim 1, wherein the first ratio is
Figure 530824DEST_PATH_IMAGE029
The second ratio is
Figure 684594DEST_PATH_IMAGE030
。
CN202211075673.4A 2022-09-05 2022-09-05 Textile flame-retardant finishing quality detection method based on vision Active CN115165886B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202211075673.4A CN115165886B (en) 2022-09-05 2022-09-05 Textile flame-retardant finishing quality detection method based on vision

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202211075673.4A CN115165886B (en) 2022-09-05 2022-09-05 Textile flame-retardant finishing quality detection method based on vision

Publications (2)

Publication Number Publication Date
CN115165886A true CN115165886A (en) 2022-10-11
CN115165886B CN115165886B (en) 2022-11-18

Family

ID=83482183

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202211075673.4A Active CN115165886B (en) 2022-09-05 2022-09-05 Textile flame-retardant finishing quality detection method based on vision

Country Status (1)

Country Link
CN (1) CN115165886B (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115689533A (en) * 2022-12-26 2023-02-03 安徽皖欣环境科技有限公司 Intelligent operation and maintenance system data management method for waste gas monitoring and treatment
CN116630899A (en) * 2023-07-21 2023-08-22 四川公路工程咨询监理有限公司 Highway side slope disease monitoring and early warning system

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH06216363A (en) * 1993-01-13 1994-08-05 Nec Corp Solid-state imaging device
US20030132847A1 (en) * 2002-01-14 2003-07-17 Anderson Kaare J. Method of detecting a fire by IR image processing
CN101634652A (en) * 2009-09-03 2010-01-27 中国烟草总公司郑州烟草研究院 Determination method for flame retardant property of tobacco tipping paper
KR20120065063A (en) * 2010-12-10 2012-06-20 한국전자통신연구원 System and method for measuring flight information of a spheric object with a high-speed stereo camera
CN103411998A (en) * 2013-08-22 2013-11-27 天津出入境检验检疫局工业产品安全技术中心 Automatic detection method for vertical combustion performance
JP2016117811A (en) * 2014-12-19 2016-06-30 三菱化学株式会社 Polycarbonate resin composition
CN112122175A (en) * 2020-08-12 2020-12-25 浙江大学 Material enhanced feature recognition and selection method of color sorter
US20210192180A1 (en) * 2018-12-05 2021-06-24 Tencent Technology (Shenzhen) Company Limited Method for training object detection model and target object detection method

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH06216363A (en) * 1993-01-13 1994-08-05 Nec Corp Solid-state imaging device
US20030132847A1 (en) * 2002-01-14 2003-07-17 Anderson Kaare J. Method of detecting a fire by IR image processing
CN101634652A (en) * 2009-09-03 2010-01-27 中国烟草总公司郑州烟草研究院 Determination method for flame retardant property of tobacco tipping paper
KR20120065063A (en) * 2010-12-10 2012-06-20 한국전자통신연구원 System and method for measuring flight information of a spheric object with a high-speed stereo camera
CN103411998A (en) * 2013-08-22 2013-11-27 天津出入境检验检疫局工业产品安全技术中心 Automatic detection method for vertical combustion performance
JP2016117811A (en) * 2014-12-19 2016-06-30 三菱化学株式会社 Polycarbonate resin composition
US20210192180A1 (en) * 2018-12-05 2021-06-24 Tencent Technology (Shenzhen) Company Limited Method for training object detection model and target object detection method
CN112122175A (en) * 2020-08-12 2020-12-25 浙江大学 Material enhanced feature recognition and selection method of color sorter

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
PING LI 等: ""Novel and eco-friendly flame-retardant cotton fabrics with lignosulfonate and chitosan through LbL: Flame retardancy, smoke suppression and flame-retardant mechanism"", 《POLYMER DEGRADATION AND STABILITY》 *
韦海成 等: ""样本熵融合聚类算法的森林火灾图像识别研究"", 《电子测量与仪器学报》 *
高震宇 等: ""基于机器视觉的卷烟阴燃速率测试系统设计"", 《传感器与微系统》 *

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115689533A (en) * 2022-12-26 2023-02-03 安徽皖欣环境科技有限公司 Intelligent operation and maintenance system data management method for waste gas monitoring and treatment
CN115689533B (en) * 2022-12-26 2023-03-07 安徽皖欣环境科技有限公司 Data management method of intelligent operation and maintenance system for waste gas monitoring and processing
CN116630899A (en) * 2023-07-21 2023-08-22 四川公路工程咨询监理有限公司 Highway side slope disease monitoring and early warning system
CN116630899B (en) * 2023-07-21 2023-10-20 四川公路工程咨询监理有限公司 Highway side slope disease monitoring and early warning system

Also Published As

Publication number Publication date
CN115165886B (en) 2022-11-18

Similar Documents

Publication Publication Date Title
CN115165886B (en) Textile flame-retardant finishing quality detection method based on vision
US5111046A (en) Apparatus and method for inspecting cooling holes
TWI603074B (en) Optical film defect detection method and system thereof
US10890537B2 (en) Appearance inspection device, lighting device, and imaging lighting device
CA2705949C (en) System and method for temperature data acquisition
US20060262971A1 (en) Transient defect detection algorithm
JP2000241362A (en) Surface quality inspecting apparatus and method
BRPI1010400A2 (en) image detection system, software, apparatus and image detection method for controlling combustion equipment
CN112730460A (en) Welding defect and intensive rosin joint detection technology for communication IC chip
CN102538000A (en) Combustion flame diagnostic method
CN111624229A (en) Intelligent charged equipment fault diagnosis method based on infrared imaging
CN114612431A (en) Visual detection method for material defects of fireproof door
CN112633292A (en) Method for measuring temperature of oxide layer on metal surface
JPH09203664A (en) Hue inspecting apparatus
KR20090057173A (en) Method for characterizing the ehaust-gas burn-off quality in combustion systems
KR101536693B1 (en) Method and system for determantion of melanoma by the multi-wavelength imaging and image processing
CN207502480U (en) Single pellet reaction measuring device
JPS60263012A (en) Flame monitoring method
CN113343791B (en) Kiln head fire-viewing video brightness identification method and device based on convolutional neural network
US6978041B2 (en) Review work supporting system
CN114663341A (en) Quantitative detection method for glare defect of optical lens
CN115077713A (en) Through hole detection method and through hole detection system
CN108986175B (en) Temperature interpretation method for temperature indicating paint area
KR20220025428A (en) Technology for real-time powder ratio measurement and moisture estimation of coke feeded from coke storage bin
KR102488651B1 (en) Apparatus for determining abnormal state based on deep learning using ir camera

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