CN113324491A - Deformation field measuring method and device based on multispectral camera - Google Patents

Deformation field measuring method and device based on multispectral camera Download PDF

Info

Publication number
CN113324491A
CN113324491A CN202110300854.1A CN202110300854A CN113324491A CN 113324491 A CN113324491 A CN 113324491A CN 202110300854 A CN202110300854 A CN 202110300854A CN 113324491 A CN113324491 A CN 113324491A
Authority
CN
China
Prior art keywords
temperature
processing module
sample
pixel
intelligent processing
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202110300854.1A
Other languages
Chinese (zh)
Inventor
唐云龙
李应朝
岳�文
康嘉杰
朱丽娜
佘丁顺
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.)
Zhengzhou Research Institute China University Of Geosciences Beijing
China University of Geosciences Beijing
Original Assignee
Zhengzhou Research Institute China University Of Geosciences Beijing
China University of Geosciences Beijing
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 Zhengzhou Research Institute China University Of Geosciences Beijing, China University of Geosciences Beijing filed Critical Zhengzhou Research Institute China University Of Geosciences Beijing
Priority to CN202110300854.1A priority Critical patent/CN113324491A/en
Publication of CN113324491A publication Critical patent/CN113324491A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
    • G01B11/00—Measuring arrangements characterised by the use of optical techniques
    • G01B11/16—Measuring arrangements characterised by the use of optical techniques for measuring the deformation in a solid, e.g. optical strain gauge

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Length Measuring Devices By Optical Means (AREA)

Abstract

The invention relates to a deformation field measuring method and device based on a multispectral camera, belonging to the technical field of material performance measurement at high temperature, and the device comprises: the device comprises an intelligent processing module, a heating module, a transmission module and an image acquisition module, wherein the acquisition of images under different channels is realized through a multispectral camera provided with a plurality of filters, a simplified brightness correction algorithm is used for correcting nonlinear brightness change caused by thermal radiation, then a corner detection method based on machine learning is adopted for carrying out corner detection on a speckle image, and finally corners in the deformed image are matched to further solve a deformation field of the speckle image under a high-temperature environment.

Description

Deformation field measuring method and device based on multispectral camera
Technical Field
The invention particularly relates to a deformation field measuring method and device based on a multispectral camera, and belongs to the technical field of material performance measurement at high temperature.
Background
The high-speed aircraft can generate a serious pneumatic heating phenomenon during supersonic flight, the physical performance and the mechanical performance of materials can be influenced by the temperature rise, for example, the strength limit and the bearing capacity of the high-speed aircraft materials can be reduced by high temperature, the geometric shape of a high-speed aircraft component can be changed, the original working state of the component can be influenced by the deviation of the geometric shape of the component at the high temperature, the generated deformation can influence the pneumatic layout of the high-speed aircraft, and further the flight safety is influenced, so that the detection of the physical performance and the mechanics of the materials in the high-temperature environment has important significance for the safety design, the reliability evaluation and the service life prediction of the materials and the structures.
The existing method for measuring deformation field mainly has contact time measurement and non-contact measurement, because the environment of measurement is in high temperature state, the environment is worse, the contact measurement method is difficult to be applied in high temperature environment, compared with the non-contact measurement method, the non-contact measurement method has the advantages of non-contact, high precision and high sensitivity, the common method is optical measurement based on image processing, but it uses the relativity of speckle images before and after deformation to carry out deformation measurement, after a certain temperature is exceeded, the relativity of speckle images is reduced, the measurement precision can not be ensured, and in the measurement, the measurement is carried out
The temperature change further causes nonlinear brightness change on the surface of the measured sample, and finally the solution accuracy of the deformation field is deteriorated.
Disclosure of Invention
The invention aims to provide a deformation field measuring method and device based on a multispectral camera, which can effectively improve the measuring precision and the measuring efficiency of a deformation field by correcting the nonlinear brightness caused by heat radiation and measuring the deformation field by using a corner point detection method based on machine learning.
A deformation field measuring method based on a multispectral camera is realized by the following steps:
the method comprises the following steps: the heating module heats the sample to be measured at high temperature, and the blue LED illuminating light source is used for irradiating the surface of the sample.
Step two: the image acquisition module is used for carrying out multi-channel image acquisition on the surface of the sample and transmitting the multi-channel image acquisition to the intelligent processing module for processing.
Step three: the intelligent processing module calculates the spectral emissivity according to the related radiation temperature measurement theory
Figure 261669DEST_PATH_IMAGE002
And the true temperature T is calculated, and then the true temperature field of the measured sample is solved.
Step four: and the intelligent processing module corrects the nonlinear brightness change caused by the heat radiation according to the real temperature field distribution.
Step five: and the intelligent processing module is used for carrying out corner detection and matching on the corrected speckle images.
Step six: meanwhile, the intelligent processing module calculates the position difference between the angular point and the matching point in the overall image subarea to obtain the final speckle image deformation field.
Further, if the image acquisition module has n channels, the voltage value of the signal output by the ith channel
Figure 490394DEST_PATH_IMAGE004
Can be expressed as
Figure 738973DEST_PATH_IMAGE006
（1）
Figure 196499DEST_PATH_IMAGE008
Is a wavelength-dependent instrument constant that is related to the spectral responsivity of the detector, the optical element transmittance, the geometry, and the first radiation constant at that wavelength,
Figure 620658DEST_PATH_IMAGE010
is the target spectral emissivity at temperature T,
Figure 911962DEST_PATH_IMAGE012
is the wavelength of the corresponding channel(s),
Figure 772471DEST_PATH_IMAGE014
is the first radiation constant and is,
Figure 282956DEST_PATH_IMAGE016
is the second radiation constant.
Further, according to the radiation thermometry theory, it is assumed that the logarithm of the emissivity and the wavelength show the following relationship,
Figure 14151DEST_PATH_IMAGE018
（2）
measuring the spectral intensity at n wavelengths to obtain n equations in accordance with formula (1), fitting the emissivity function and dynamic temperature by using regression algorithm, and solving to obtain spectral emissivity
Figure 476357DEST_PATH_IMAGE010
And after the real temperature T is obtained, the real temperature of each unit can be calculated by reading the current of each unit of the multispectral camera, and then a real temperature field can be solved.
Further, an intelligent processing module obtains the pixel gray level number of each corrected brightness area through an algorithm to correct the brightness nonlinear change caused by heat radiation, and the specific algorithm implementation method is as follows:
s1: the intelligent processing module equally divides the acquired speckle images by n, and calculates the pixel brightness value of each area
Figure 574894DEST_PATH_IMAGE020
。
S2: obtaining a plurality of intervals according to the temperature field distribution of the speckle image and the pixel brightness value of each area
Figure 780747DEST_PATH_IMAGE022
，
Figure 366449DEST_PATH_IMAGE024
，···，
Figure 842299DEST_PATH_IMAGE026
Fitting the values of the regions to obtain the pixel brightness values of the regions
Figure 146241DEST_PATH_IMAGE028
And temperature
Figure 765573DEST_PATH_IMAGE030
The relationship between
Figure 471360DEST_PATH_IMAGE032
，
Figure 540948DEST_PATH_IMAGE034
，···，
Figure 581454DEST_PATH_IMAGE036
In order to make the corrected error smaller, the correlation coefficient of the mathematical function expression linearly fitted to each region should be maximized when fitting to each interval.
S3: solving the inverse function of each curve equation separately
Figure 722585DEST_PATH_IMAGE038
。
S4: gray level division: the gray level section of the speckle image to be collected according to the number m of gray levels to be displayed
Figure 768032DEST_PATH_IMAGE040
，
Figure 398734DEST_PATH_IMAGE042
]M-1 parts by weight, and a brightness corresponding to a gray level K of
Figure 660957DEST_PATH_IMAGE044
To judge the brightness
Figure 340200DEST_PATH_IMAGE044
In which section
Figure 161525DEST_PATH_IMAGE046
Dividing the corresponding gray level K into the segment, and calculating the temperature corresponding to each gray level K according to the inverse function in step S3
Figure 713861DEST_PATH_IMAGE048
，
Figure 620637DEST_PATH_IMAGE050
。
S5: counting the brightness values of all pixels at the same temperature, and taking the intermediate brightness value
Figure 149576DEST_PATH_IMAGE052
Comparing the brightness values of two pixel values with the same temperature value when
Figure 215621DEST_PATH_IMAGE054
When, if
Figure 63491DEST_PATH_IMAGE056
And
Figure 332930DEST_PATH_IMAGE058
let us order
Figure 26079DEST_PATH_IMAGE060
I.e. by
Figure DEST_PATH_IMAGE061
And
Figure 149893DEST_PATH_IMAGE028
the pixel gray level of the region is adjusted to
Figure 276987DEST_PATH_IMAGE062
Corresponding gray scale
Figure 158355DEST_PATH_IMAGE064
When is coming into contact with
Figure 779829DEST_PATH_IMAGE066
When the temperature of the water is higher than the set temperature,
Figure 430253DEST_PATH_IMAGE028
taking the maximum brightness value of the region
Figure 964134DEST_PATH_IMAGE061
I.e. by
Figure 332799DEST_PATH_IMAGE061
And
Figure 492384DEST_PATH_IMAGE028
the pixel gray level of the region is adjusted to
Figure 262894DEST_PATH_IMAGE061
The corresponding gray level K.
Further, the intelligent processing module performs corner detection on each speckle image by using a machine learning-based corner detection method, and the specific algorithm implementation flow is as follows:
scheme 1: and (4) training the corrected speckle images, and detecting characteristic points of each speckle image by using a corner detection algorithm.
And (2) a flow scheme: for each pixel P, the 16 pixels surrounding it are saved as a vector P, and step A, B is repeated for all pixels in the speckle image.
And (3) a flow path: for each value in the vector P there are 3 states, darker than the pixel P and of a quantity similar to the pixel P, the pixel P being divided into 3 subsets according to its state: pd, Pl, Ps.
And (4) a flow chart: and defining a corresponding variable Kp for each vector, wherein the KP is true when P is a characteristic point, and the Kp is false when P is not the characteristic point.
And (5) a flow chart: the least quotient method is recursively applied to the three subsets Pd, Pl, Ps until it stops when the subset is 0.
Furthermore, the intelligent processing module performs feature point matching on the deformed image to obtain matching points of the corner points, calculates the position difference between the corner points of all sub-regions in the speckle image and the matching points, and takes the average value as the displacement value of the center point of the sub-region of the image to further obtain the deformed field of the speckle image.
Further, a certain pixel point (X) in the speckle image is used0,Y0) As a center, the speckle image before deformation is divided into image sub-regions of (2M +1) pixel size, where M represents a half of the length of the sub-region.
Further, the deformation field measuring device based on the multispectral camera comprises a heating module, an image acquisition module, an intelligent processing module and a transmission module.
And the signal input ends of the heating module, the image acquisition module and the transmission module are connected with the signal output end of the intelligent processing module.
The image acquisition module comprises an infrared temperature measurement sensor, a temperature sensor, a blue LED illumination light source, a multispectral camera and a liquid cooling device.
The transmission module comprises an X-direction transmission lead screw, a lifting disc and a Y-direction transmission lead screw.
Furthermore, the heating module heats the surface of the sample, the infrared temperature measuring sensor measures the central temperature of the surface of the sample, the central temperature is transmitted to the intelligent processing module in real time, and the working state of the heating module is remotely controlled by a worker through the intelligent processing module.
According to the front view of the device, an experiment inlet and an experiment outlet are arranged right above the device, the transmission module comprises a feed motion in X, Y and Z directions, the transmission process in the X direction is that a lead screw transmits and retreats the lifting disc, the transmission process in the Y direction is that a clamp clamps and loosens a sample through the motion of the lead screw, and the motion in the Z direction is that the lifting disc feeds and retreats the sample through the up-and-down motion.
The multispectral camera carries out image acquisition on a sample and transmits the image acquisition to the intelligent processing module for processing in real time, the blue LED lighting source carries out light supplement on the surface of the sample in real time, the temperature sensor measures the ambient temperature around the multispectral camera and transmits temperature data to the intelligent processing module for alarm prompt, and the liquid cooling device cools the ambient environment around the multispectral camera.
Furthermore, the automatic lifting disc comprises a conveying clamp, and the conveying clamp consists of a clamping block and a small spring. The intelligent processing module sends the sample to the lower part of the Y-direction screw clamp through controlling the screw rod of the X-direction transmission and the lifting disc to be matched for transmission, meanwhile, the sample is clamped through controlling the movement of the Y-direction screw rod, after the measurement work is completed, the intelligent processing module conducts loosening clamp through controlling the Y-direction screw rod to be reversely transmitted, the sample falls into the lifting disc, and the sample is sent to the outlet of the measuring device through the X-direction screw rod and the lifting disc to be matched for transmission.
By adopting the technical scheme, the invention has the advantages that:
1. the invention uses a simplified nonlinear brightness correction algorithm, and aims at the problem that the same temperature area has different brightness in the collected speckle image, the correction algorithm divides the speckle image into n areas, fits the brightness and temperature of each area by combining the temperature field distribution of the whole speckle image, further determines the function relation of the temperature and the brightness of each area, divides the global gray level of the speckle image, the different gray levels correspond to the different brightness levels, and finally determines the gray level of each area by comparing the brightness values of the same temperature area, thereby achieving the purpose of correcting the nonlinear brightness caused by heat radiation, and effectively improving the precision of deformation field measurement.
2. According to the method, the angular point of the speckle image is detected by using the angular point detection method based on machine learning, when the detection is fast, if a group of n adjacent pixels (n of 16 pixels) in a circle is smaller than 12, excessive characteristic points are prevented from being detected, the accuracy of the method does not depend on the position and distribution of the characteristic points, the optimal characteristic points can be selected, and the detection accuracy of the characteristic points is effectively improved.
3. The intelligent deformation field measuring device is used for realizing automatic feeding and material returning in the sample measuring process, avoiding the damage of a high-temperature environment in a closed cavity to a human body after the measurement is finished and effectively improving the deformation field measuring efficiency.
Drawings
Fig. 1 is a simplified flow chart of non-linear luminance correction.
Fig. 2 is a schematic diagram of corner detection.
Fig. 3 is a flow chart of corner detection based on machine learning.
Fig. 4 is a speckle image measurement area diagram.
Fig. 5 is a front view of a deformation field measuring device based on a multispectral camera.
Fig. 6 is a top view of a deformation field measuring device based on a multispectral camera.
Reference numerals: 1-heating module, 2-sample, 3-image acquisition module, 4-intelligent processing module, 5-lifting plate, 6-conveying clamp, 7-sample clamp and 8-lead screw.
Detailed Description
The following examples are intended to illustrate the invention but are not intended to limit the scope of the invention.
In the embodiment, the sample 2 is a silicon carbide material, the cross section is rectangular, speckles are formed by manually and uniformly spraying the speckles on the surface of the sample 2 to be measured at random, and the deformation field measuring method and the device based on the multispectral camera have the following specific implementation steps:
the method comprises the following steps:
place sample 2 in the conveying anchor clamps 6 of lifter plate, by the motion of intelligent processing module 4 control X direction lead screw and Z direction lifter plate 5 to deliver to Y direction lead screw department anchor clamps 7 below, Y direction lead screw motion drives anchor clamps 7 motion, and then realizes pressing from both sides the tight of material.
Step two:
the method comprises the steps of continuously heating one side of a sample 2 to be measured at high temperature by using oxyacetylene as a heating module 1 in a measuring device, wherein the heating interval is 100-1000 ℃, meanwhile, a blue LED light source is used for irradiating the surface of a material, and meanwhile, an infrared temperature sensor and a temperature sensor measure the surface temperature of the sample 2 and the temperature of an image acquisition module 3 and transmit the temperature to an intelligent processing module 4 in real time.
Step three:
the surface of the sample 2 is subjected to multi-channel image acquisition at a certain temperature through the image acquisition module 3 and is transmitted to the intelligent processing module 4 for processing.
Step four:
if the image acquisition module 3 has n channels, the voltage value of the signal output by the ith channel
Figure 466211DEST_PATH_IMAGE004
Can be expressed as
Figure 322172DEST_PATH_IMAGE006
（1）
Figure 285449DEST_PATH_IMAGE008
Is a wavelength-dependent instrument constant that is related to the spectral responsivity of the detector, the optical element transmittance, the geometry, and the first radiation constant at that wavelength,
Figure 644886DEST_PATH_IMAGE010
is the target spectral emissivity at the temperature T,
Figure 786148DEST_PATH_IMAGE012
is the wavelength of the corresponding channel(s),
Figure 863826DEST_PATH_IMAGE014
is the first radiation constant and is,
Figure 630794DEST_PATH_IMAGE016
is the second radiation constant.
Taking logarithm of (1) type two ends respectively to obtain formula (2)
Figure 110316DEST_PATH_IMAGE068
（2）
Order to
Figure 124278DEST_PATH_IMAGE070
When the formula (2) is changed to
Figure 282727DEST_PATH_IMAGE072
（3）
According to the radiation thermometry theory, it is assumed that the logarithm of the emissivity and the wavelength show the following relationship,
Figure DEST_PATH_IMAGE073
substituting into (3) and multiplying by λ at both ends to obtain formula (4)
Figure DEST_PATH_IMAGE075
（4）
Measuring the spectral intensity at n wavelengths to obtain n equations in accordance with formula (1), fitting the emissivity function and dynamic temperature by using regression algorithm, and solving to obtain spectral emissivity
Figure 712440DEST_PATH_IMAGE010
After the real temperature T is reached, the real temperature of each unit can be calculated by reading the current of each unit of the image acquisition module 3, and then a real temperature field can be solved.
Step five:
as shown in fig. 1, in a simplified non-linear luminance correction method, the intelligent processing module 4 obtains the pixel gray scale number of each modified region through an algorithm to correct the non-linear luminance variation caused by heat radiation, and the specific algorithm implementation method (S1-S5) is as follows:
s1: the intelligent processing module 4 equally divides the acquired speckle images by n, and calculates the pixel brightness value of each area
Figure 46469DEST_PATH_IMAGE020
。
S2: obtaining a plurality of intervals according to the temperature field distribution of the speckle image and the brightness value of each pixel of each area
Figure 778802DEST_PATH_IMAGE076
，
Figure DEST_PATH_IMAGE077
，⋯，
Figure 909700DEST_PATH_IMAGE078
Fitting the values of the regions to obtain the pixel brightness values of the regions
Figure DEST_PATH_IMAGE079
And temperature
Figure 64476DEST_PATH_IMAGE028
The relationship between
Figure 253011DEST_PATH_IMAGE080
，
Figure DEST_PATH_IMAGE081
，···，
Figure DEST_PATH_IMAGE082
In order to make the corrected error smaller, the correlation coefficient of the mathematical function expression linearly fitted to each region should be maximized when fitting to each interval.
S3: solving the inverse function of each curve equation separately
Figure 31612DEST_PATH_IMAGE038
。
S4: gray level division: the gray level section of the speckle image to be collected according to the number m of gray levels to be displayed
Figure 571177DEST_PATH_IMAGE040
，
Figure 467327DEST_PATH_IMAGE042
]M-1 parts by weight, and a brightness corresponding to a gray level K of
Figure 775948DEST_PATH_IMAGE044
To judge the brightness
Figure 850084DEST_PATH_IMAGE044
In which section
Figure 611366DEST_PATH_IMAGE046
Dividing the corresponding gray level K into the segment, and calculating the temperature corresponding to each gray level K according to the inverse function in step S3
Figure 812672DEST_PATH_IMAGE048
，
Figure 975800DEST_PATH_IMAGE050
。
S5: counting the brightness values of all pixels at the same temperature, and taking the intermediate brightness value
Figure 220836DEST_PATH_IMAGE052
Comparing the brightness values of two pixel values with the same temperature value when
Figure 531732DEST_PATH_IMAGE054
When, if
Figure 395783DEST_PATH_IMAGE056
And
Figure 623256DEST_PATH_IMAGE058
let us order
Figure 180139DEST_PATH_IMAGE060
I.e. by
Figure 40648DEST_PATH_IMAGE061
And
Figure 442811DEST_PATH_IMAGE028
region(s)Is adjusted to
Figure 190318DEST_PATH_IMAGE062
Corresponding gray scale
Figure 918102DEST_PATH_IMAGE064
When is coming into contact with
Figure 265907DEST_PATH_IMAGE066
When the temperature of the water is higher than the set temperature,
Figure 471761DEST_PATH_IMAGE028
taking the maximum brightness value of the region
Figure 260725DEST_PATH_IMAGE061
I.e. by
Figure 533312DEST_PATH_IMAGE061
And
Figure 978200DEST_PATH_IMAGE028
the pixel gray level of the region is adjusted to
Figure 112378DEST_PATH_IMAGE061
Corresponding gray scale
Figure DEST_PATH_IMAGE084
。
Step six:
as shown in fig. 2-3, the intelligent processing module 4 performs feature point detection on each speckle image by using a machine learning-based corner point detection method, and the specific algorithm implementation flow is as follows:
scheme 1: and (4) training the corrected speckle images, and detecting characteristic points of each speckle image by using a corner detection algorithm.
And (2) a flow scheme: a pixel P is selected in the corrected picture with a pixel intensity IP.
And (3) a flow path: a threshold T is set, typically 20% of the picture pixel intensity.
And (4) a flow chart: with the pixel point P as the center, there are 16 pixel points (P1, P2.., P16) on the circle with radius 3, and the pixel differences between these pixel points and the center point P are calculated respectively.
And (5) a flow chart: and if the absolute value of the pixel difference of n continuous pixel points is greater than the threshold value T, the central point P is a characteristic point.
And (6) a flow path: to speed up the algorithm, only the intensity differences of the pixel points p1, p5, p9, p13 and the center point are compared initially. If the pixel difference of at least 3 points is greater than the threshold, then P is a feature candidate point.
Scheme 7: and (3) performing the step (3) on the characteristic candidate points, judging all 16 pixel points, and determining whether p is a characteristic point.
And (3) a process 8: repeating the steps A2-A8 for all the pixel points.
And (3) a process 9: as shown in FIG. 2, for each pixel P, the 16 pixels surrounding it are saved as a vector P, and the steps A1-A9 are repeated for all pixels in the speckle image.
A process 10: for each value in the vector P there are 3 states, darker than the pixel P and of a quantity similar to the pixel P, the pixel P being divided into 3 subsets according to its state: pd, Pl, Ps,
Figure DEST_PATH_IMAGE086
and defining a corresponding variable Kp for each vector, wherein the KP is true when P is a characteristic point, and the Kp is false when P is not the characteristic point.
Scheme 11: querying each subset using a decision tree classifier, and using the variable Kp as a classification value of the training set,
Figure DEST_PATH_IMAGE088
。
and (3) a process 12: the least quotient method is recursively applied to the three subsets Pd, Pl, Ps until it stops when the subset is 0.
Step seven:
as shown in FIG. 4, the intelligent processing module 4 will transform the pre-distorted imageThe area with measurement is divided into grid form, the node distance between grids is 3-5 pixels, and in the specific calculation of deformation measurement, one point (X) is extracted from the image before deformation for a certain grid point0，Y0) The speckle image is an image subarea with the pixel size of (2M +1) (2M +1) as the center, the intelligent processing module 4 tracks and matches the speckle image before and after deformation by using a characteristic point matching algorithm to obtain matching points of angular points, then the position difference between the angular points and the matching points of the image subarea is calculated, the average value is taken as the center point of the image subarea, and the deformation field of the speckle image is finally solved through the displacement value of the grid points.
Step eight:
after the deformation field measurement work is finished, the intelligent processing module controls the Y-direction lead screw to perform reverse transmission to loosen the clamp, the sample falls into the lifting disc, and the X-direction lead screw and the lifting disc are matched to transmit the sample to the outlet of the measuring device.
As shown in fig. 5-6, a deformation field measuring device with brightness correction based on a multispectral camera comprises a heating module 1, a switch signal control end of the heating module 1 is connected with an intelligent processing module 3, an image acquisition module 3 comprises an infrared temperature measurement sensor, a temperature sensor, a blue LED illumination light source, the multispectral camera and a liquid cooling device, signal output ends of the infrared temperature measurement sensor and the temperature sensor are connected with a signal input end of the intelligent processing module, the blue LED illumination light source is arranged right above a lens of the multispectral camera and used for supplementing light to the surface of a sample, the infrared temperature measurement sensor is arranged right below the lens of the multispectral camera, a signal output end of the multispectral camera is connected with a signal input end of the intelligent processing module 4, a lifting disc 5 is fixed on a supporting plate connected with a transmission screw in the X direction, the lifting disc moves up and down through a motor fixed on the supporting plate, the conveying clamp 6 comprises clamping blocks and springs, the two clamping blocks are connected through the springs, the two clamping blocks freely slide in the Y direction in a groove embedded in the lifting disc 5, and then clamping and loosening of the sample 2 are achieved.
The transmission module comprises an X-direction transmission lead screw 8, a lifting disc 5 and a clamp 7 on the Y-direction transmission lead screw.
The heating module 1 heats the surface of the sample 2, the infrared temperature measuring sensor measures the central temperature of the surface of the sample 2, the central temperature is transmitted to the intelligent processing module 4 in real time, and a worker remotely controls the working state of the heating module 1 through the intelligent processing module 4.
From the front view of the device, an experiment inlet and an experiment outlet are arranged right above the device, the transmission module has feeding motion in the X direction, the Y direction and the Z direction, the transmission process in the X direction is the transmission and the return of the screw rod to the lifting disc 5, the transmission process in the Y direction is the clamping and the loosening of the clamp 7 to the sample 2 through the screw rod motion, and the motion in the Z direction is the feeding and the returning of the lifting disc 5 to the sample 2 through the up-and-down motion.
The image acquisition module 3 carries out image acquisition to sample 2 to real-time transmission handles to intelligent processing module 4, and blue LED illuminating light source carries out the light filling to sample 2 surface in real time, and temperature sensor measures multispectral camera surrounding environment temperature, and sends temperature data to intelligent processing module 4, and the liquid cooling device lasts the cooling to multispectral camera surrounding environment.
Although the invention has been described in detail above with reference to a general description and specific examples, it will be apparent to one skilled in the art that modifications or improvements may be made thereto based on the invention. Accordingly, such modifications and improvements are intended to be within the scope of the invention as claimed.

Claims (10)

1. A deformation field measuring method based on a multispectral camera is characterized by comprising the following steps:
the method comprises the following steps: the heating module is used for heating the sample to be measured at high temperature, and the blue LED illumination light source is used for irradiating the surface of the sample;
step two: the image acquisition module acquires multi-channel images of the surface of the sample and transmits the images to the intelligent processing module for processing;
step three: the intelligent processing module calculates the spectral emissivity according to the related radiation temperature measurement theory
Figure 884335DEST_PATH_IMAGE002
And the true temperature T, and then solving the true temperature field of the measured sample;
step four: the intelligent processing module corrects the nonlinear brightness change caused by the heat radiation according to the real temperature field distribution;
step five: the intelligent processing module carries out angular point detection and matching on the corrected speckle images;
step six: meanwhile, the intelligent processing module calculates the position difference between the angular point and the matching point in the overall image subarea to obtain the final speckle image deformation field.
2. The method according to claim 1, wherein the deformation field measurement method based on the multispectral camera comprises the following steps:
if the image acquisition module has n channels, the voltage value of the signal output by the ith channel
Figure 988426DEST_PATH_IMAGE004
Can be expressed as
Figure 705846DEST_PATH_IMAGE006
（1）
Figure 553585DEST_PATH_IMAGE008
Is a wavelength-dependent instrument constant that is related to the spectral responsivity of the detector, the optical element transmittance, the geometry, and the first radiation constant at that wavelength,
Figure 305641DEST_PATH_IMAGE010
is the target spectral emissivity at the temperature T,
Figure 315054DEST_PATH_IMAGE012
is the wavelength of the corresponding channel(s),
Figure 519770DEST_PATH_IMAGE014
is the first radiation constant and is,
Figure 171200DEST_PATH_IMAGE016
is the second radiation constant.
3. The method according to claim 1, wherein the deformation field measurement method based on the multispectral camera comprises the following steps:
according to the radiation thermometry theory, it is assumed that the logarithm of the emissivity and the wavelength show the following relationship,
Figure 777762DEST_PATH_IMAGE018
（2）
measuring the spectral intensity at n wavelengths to obtain n equations in accordance with formula (1), fitting the emissivity function and dynamic temperature by using regression algorithm, and solving to obtain spectral emissivity
Figure 958077DEST_PATH_IMAGE010
And after the real temperature T is obtained, the real temperature of each unit can be calculated by reading the current of each unit of the multispectral camera, and then a real temperature field can be solved.
4. The method according to claim 1, wherein the deformation field measurement method based on the multispectral camera comprises the following steps: a simplified non-linear brightness correction method, intelligent processing module corrects the non-linear brightness change caused by heat radiation by obtaining the pixel gray scale number of each brightness area, the specific algorithm implementation method is as follows:
s1: the intelligent processing module equally divides the acquired speckle images by n, and calculates the pixel brightness value of each area
Figure 181248DEST_PATH_IMAGE020
；
S2: obtaining a plurality of intervals according to the temperature field distribution of the speckle image and the brightness value of each pixel of each area
Figure 855942DEST_PATH_IMAGE022
，
Figure 566278DEST_PATH_IMAGE024
，···，
Figure 199385DEST_PATH_IMAGE026
Fitting the values of the regions to obtain the pixel brightness values of the regions
Figure 627961DEST_PATH_IMAGE028
And temperature
Figure 840768DEST_PATH_IMAGE030
The relationship between
Figure 936769DEST_PATH_IMAGE032
，
Figure 944039DEST_PATH_IMAGE034
，···，
Figure 125490DEST_PATH_IMAGE036
In order to make the corrected error smaller, the correlation coefficient of the mathematical function expression linearly fitted in each region is made the largest when fitting in each interval;
s3: solving the inverse function of each curve equation separately
Figure 141988DEST_PATH_IMAGE038
；
S4: gray level division: the gray level section of the speckle image to be collected according to the number m of gray levels to be displayed
Figure 843228DEST_PATH_IMAGE040
，
Figure 801825DEST_PATH_IMAGE042
]M-1 parts by weight, and a brightness corresponding to a gray level K of
Figure 690147DEST_PATH_IMAGE044
To judge the brightness
Figure 494024DEST_PATH_IMAGE044
In which section
Figure 315349DEST_PATH_IMAGE046
The corresponding gray level K is divided into the section, and then each gray level is calculated according to the inverse function in step S3
Figure 710427DEST_PATH_IMAGE048
Corresponding temperature
Figure 86045DEST_PATH_IMAGE050
，
Figure 959192DEST_PATH_IMAGE052
；
S5: counting the brightness values of all pixels at the same temperature, and taking the intermediate brightness value
Figure 635024DEST_PATH_IMAGE054
Comparing the brightness values of two pixel values with the same temperature value when
Figure 686157DEST_PATH_IMAGE056
When, if
Figure 63917DEST_PATH_IMAGE058
And
Figure 225908DEST_PATH_IMAGE060
let us order
Figure 739935DEST_PATH_IMAGE062
I.e. by
Figure 961969DEST_PATH_IMAGE044
And
Figure 843337DEST_PATH_IMAGE028
the pixel gray level of the region is adjusted to
Figure DEST_PATH_IMAGE063
Corresponding gray scale
Figure DEST_PATH_IMAGE065
When is coming into contact with
Figure DEST_PATH_IMAGE067
When the temperature of the water is higher than the set temperature,
Figure 182921DEST_PATH_IMAGE028
taking the maximum brightness value of the region
Figure 833345DEST_PATH_IMAGE044
I.e. by
Figure 491859DEST_PATH_IMAGE044
And
Figure 578633DEST_PATH_IMAGE028
the pixel gray level of the region is adjusted to
Figure 613585DEST_PATH_IMAGE044
The corresponding gray level K.
5. The method according to claim 1, wherein the deformation field measurement method based on the multispectral camera comprises the following steps: the intelligent processing module uses a machine learning-based corner detection method to perform corner detection on each speckle image, and the specific algorithm implementation flow is as follows:
scheme 1: using the corrected speckle images for training, and detecting characteristic points of each speckle image by using a corner detection algorithm;
and (2) a flow scheme: for each pixel point P, saving 16 pixel points surrounding it as a vector P, and repeating step A, B for all pixels in the speckle image;
and (3) a flow path: for each value in the vector P there are 3 states, darker than the pixel P and of a quantity similar to the pixel P, the pixel P being divided into 3 subsets according to its state: pd, Pl, Ps;
and (4) a flow chart: defining a corresponding variable Kp for each vector, wherein when P is a characteristic point, KP is true, and when P is not a characteristic point, Kp is false;
and (5) a flow chart: querying each subset by using a decision tree classifier, and taking a variable Kp as a classification value of a training set;
and (6) a flow path: the least quotient method is recursively applied to the three subsets Pd, Pl, Ps until it stops when the subset is 0.
6. The method according to claim 1, wherein the deformation field measurement method based on the multispectral camera comprises the following steps:
the intelligent processing module performs characteristic point matching on the deformed image to obtain matching points of the angular points, calculates the position difference between the angular points of all sub-areas in the speckle image and the matching points, and takes the average value as the displacement value of the central point of the sub-area of the image to further obtain the deformed field of the speckle image.
7. The method according to claim 7, wherein the deformation field measurement method based on the multispectral camera comprises the following steps:
and (3) dividing the image subareas of the speckle images before deformation in the pixel size of (2M +1) (2M +1) by taking a certain pixel point (X0, Y0) in the speckle images as a center, wherein M represents the length of half of the subareas.
8. A deformation field measuring device based on a multispectral camera is characterized in that:
the measuring device comprises a heating module, an image acquisition module, an intelligent processing module and a transmission module;
the signal input ends of the heating module, the image acquisition module and the transmission module are connected with the signal output end of the intelligent processing module;
the image acquisition module comprises an infrared temperature measurement sensor, a temperature sensor, a blue LED illumination light source, a multispectral camera and a liquid cooling device;
the transmission module comprises an X-direction transmission lead screw, a lifting disc and a Y-direction transmission lead screw.
9. The multi-spectral camera based deformation field measuring device according to claim 8, wherein:
the heating module heats the surface of the sample, the infrared temperature measuring sensor measures the central temperature of the surface of the sample, the central temperature is transmitted to the intelligent processing module in real time, and a worker remotely controls the working state of the heating module through the intelligent processing module;
the front view of the device shows that an experiment inlet and an experiment outlet are arranged right above the device, the transmission module comprises feeding motion in X, Y and Z directions, the transmission process in the X direction is that a lead screw transmits and retreats the lifting disc, the transmission process in the Y direction is that a clamp clamps and looses a sample through the movement of the lead screw, and the transmission process in the Z direction is that the lifting disc feeds and retreats the sample through up-and-down movement;
the multispectral camera carries out image acquisition on a sample and transmits the image acquisition to the intelligent processing module for processing in real time, the blue LED illuminating light source carries out real-time light supplement on the surface of the sample, the temperature sensor measures the ambient temperature around the multispectral camera and transmits temperature data to the intelligent processing module for alarm prompt, and the liquid cooling device cools the ambient environment around the multispectral camera.
10. The multi-spectral camera based deformation field measuring device according to claim 8, wherein:
the lifting disc comprises a conveying clamp, and the conveying clamp consists of a clamping block and a spring. The intelligent processing module is used for controlling the X-direction transmission screw rod and the lifting disc to be in matched transmission and sending a sample to the lower part of the Y-direction screw rod clamp, clamping the sample by controlling the motion of the Y-direction screw rod, after the measurement work is completed, loosening the clamp by controlling the Y-direction screw rod to be in reverse transmission, dropping the sample in the lifting disc, and sending the sample to the outlet of the measuring device through the X-direction transmission screw rod and the lifting disc in matched transmission.
CN202110300854.1A 2021-03-22 2021-03-22 Deformation field measuring method and device based on multispectral camera Pending CN113324491A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110300854.1A CN113324491A (en) 2021-03-22 2021-03-22 Deformation field measuring method and device based on multispectral camera

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110300854.1A CN113324491A (en) 2021-03-22 2021-03-22 Deformation field measuring method and device based on multispectral camera

Publications (1)

Publication Number Publication Date
CN113324491A true CN113324491A (en) 2021-08-31

Family

ID=77414763

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110300854.1A Pending CN113324491A (en) 2021-03-22 2021-03-22 Deformation field measuring method and device based on multispectral camera

Country Status (1)

Country Link
CN (1) CN113324491A (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114754694A (en) * 2022-06-13 2022-07-15 新乡职业技术学院 Material deformation detection equipment based on it is multispectral
CN116723409A (en) * 2022-02-28 2023-09-08 荣耀终端有限公司 Automatic exposure method and electronic equipment
CN116723409B (en) * 2022-02-28 2024-05-24 荣耀终端有限公司 Automatic exposure method and electronic equipment

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116723409A (en) * 2022-02-28 2023-09-08 荣耀终端有限公司 Automatic exposure method and electronic equipment
CN116723409B (en) * 2022-02-28 2024-05-24 荣耀终端有限公司 Automatic exposure method and electronic equipment
CN114754694A (en) * 2022-06-13 2022-07-15 新乡职业技术学院 Material deformation detection equipment based on it is multispectral

Similar Documents

Publication Publication Date Title
EP3267185B1 (en) Method for turbine component qualification
CN103471910B (en) A kind of elongation at break of metal material intelligent test method followed the tracks of based on random point
Jia et al. An improved image acquiring method for machine vision measurement of hot formed parts
US9113046B2 (en) Tire contour measurement data correction method and tire visual inspection device
CN113324491A (en) Deformation field measuring method and device based on multispectral camera
CN1141557C (en) Image detection method of combustion temperature in hearth
JP6777428B2 (en) Temperature measuring device
CN110073008A (en) Blast furnace iron notch temperature measuring device
US9415464B2 (en) Laser machining system utilizing thermal radiation image and method thereof
CN110160661B (en) Object surface temperature measuring method and device based on visible light photo
CN111707382A (en) Dynamic optical compensation method and device for synchronous measurement of temperature deformation
TW201321728A (en) Non-contact temperature measuring method
US9664568B2 (en) Extended temperature mapping process of a furnace enclosure with multi-spectral image-capturing device
CN117146733A (en) Comprehensive measurement method for high Wen Mubiao three-dimensional morphology and temperature field
GB2596009A (en) Night vision device testing
CN116067506A (en) Method and system for constructing three-dimensional temperature field on material surface of industrial furnace
Wang et al. A temperature measurement method based on visible light chromaticity index and k-nearest neighbors algorithm
JP6570059B2 (en) Non-contact temperature measurement method and measurement system
Dauvin et al. Optimization of temperature, targets, and illumination for high precision photogrammetric measurements
CN107631798A (en) The wavelength scaling method and system of push-broom type imaging spectrometer
JP2019020238A (en) Non-contact external shape measuring system of red-hot press molding material by means edge measurement
WO2019159204A1 (en) Device for measuring temperature
JP6015084B2 (en) Method and apparatus for measuring length of long body
Zhang et al. Research and Application of Temperature Monitoring and Error Correction System for High-Strength Steel Hot Stamping Parts
CN113654667B (en) Infrared radiation temperature measurement method for reducing environment background radiation noise

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication