An infrared measurement method and device for circumferential temperature field distribution of an elongated object

Abstract

The application discloses a kind of circumferential temperature field distribution thermometry methods and devices for elongated object.It uses visual sensor to collect the image information of measured elongated object, determines the three-dimensional space position and deformation information of measured elongated object by image processing algorithm, and according to the position and deformation information, drives motion platform to carry out dynamic adjustment;Then utilize multiple infrared sensors to surround the circumferential distribution of measured elongated object, synchronously collect its surface temperature data;Finally, the collected temperature data is corrected, the data of multiple infrared sensors is fused, the circumferential continuous temperature distribution of the surface of measured elongated object is reconstructed, and the complete two-dimensional temperature field distribution is generated combined with axial scanning data.The application solves the problems such as discrete sampling, contact interference and poor deformation adaptability in traditional technology by non-contact measurement, realizes the continuous high-precision measurement of circumferential-axial two-dimensional temperature field of elongated object surface.

Description

Technical Field

[0001] This invention relates to the field of infrared measurement of object surfaces, and specifically to a method and apparatus for measuring the circumferential temperature field distribution of slender objects. Background Technology

[0002] As a special functional fuel for hypersonic vehicles, endothermic fuels must couple the functions of cooling and heat absorption with combustion and propulsion within the cooling channel. Accurately obtaining their thermophysical parameters and characterizing their thermochemical behavior requires various simulated heat transfer experimental devices and high-precision calorimetric systems. Among these, achieving high-precision measurement of the wall temperature of a tubular calorimeter is particularly crucial, predicated on the ability to accurately measure the two-dimensional temperature field distribution along the circumferential direction of the tube wall surface. This measurement provides core data support for studying the heat transfer process between the fluid and the tube wall, as well as for determining the fluid's own thermophysical parameters.

[0003] However, continuous heat exchange occurs between the calorimeter surface and the fluid, and due to the complex heat absorption / release characteristics of the fluid, a significant temperature gradient forms on the tube wall surface. These gradients exhibit high spatial complexity in both the axial and circumferential directions, making it difficult for traditional temperature measurement methods to meet the high-precision measurement requirements of temperature fields in such slender structures.

[0004] (1) Contact temperature measurement is generally achieved by arranging several thermocouples at equal intervals along the axial direction of the object being measured, or by using other fixed-point non-contact temperature measurement methods. However, such fixed temperature measurement methods can often only obtain the temperature of a few discrete points and cannot obtain the continuous circumferential temperature distribution of the object. At the same time, contact temperature measurement often destroys the original temperature field of the pipe being measured, making it impossible to set up a large number of temperature measurement points.

[0005] (2) A one-dimensional displacement stage based on an infrared thermometer can be built to achieve one-dimensional scanning of a straight line, but this method cannot scan irregular objects or objects that undergo dynamic deformation, and its applicability is poor.

[0006] (3) Using multiple thermal imagers can theoretically achieve temperature measurement by stitching together the images, but the cost is high.

[0007] (4) By combining a thermal imager and a point thermometer to track and scan the pipe under test on a two-dimensional motion stage, the axial temperature distribution of the pipe under test can also be obtained. However, this method assumes that the circumferential temperature distribution of the pipe under test is uniform. When the circumferential temperature distribution on the outer surface of the object is not uniform, this measurement will also cause certain errors.

[0008] Therefore, the infrared measurement method and device for circumferential temperature field distribution of slender objects proposed in this invention can effectively solve the problem that traditional methods cannot effectively measure slender objects with complex temperature gradients, providing a new solution for high-precision temperature field monitoring and providing reliable data support for related application fields. Summary of the Invention

[0009] This invention addresses the need for measuring the circumferential temperature field distribution of slender objects by proposing an infrared temperature measurement method and device based on multi-sensor collaboration, three-dimensional dynamic tracking, and image fusion.

[0010] The present invention provides an infrared measurement method for circumferential temperature field distribution of a slender object, comprising the following steps:

[0011] Dynamic tracking and positioning: At least two vision sensors are used to acquire image information of the slender object being measured. The three-dimensional spatial position and deformation information of the slender object being measured are determined by the image processing algorithm. The motion platform is driven to make dynamic adjustments based on the position and deformation information so that the measuring device and the slender object being measured maintain a relatively stable positional relationship.

[0012] Temperature data acquisition: Multiple infrared sensors are distributed around the circumference of the slender object being measured to simultaneously collect temperature data of its surface.

[0013] Data processing and temperature field reconstruction: The collected temperature data is corrected and then the data from multiple infrared sensors are fused using a data fusion algorithm to reconstruct the circumferential continuous temperature distribution on the surface of the slender object being measured, and a complete two-dimensional temperature field distribution is generated by combining the axial scan data.

[0014] This invention solves the problems of discrete sampling, contact interference and poor deformation adaptability in traditional technologies by non-contact measurement, and realizes continuous high-precision measurement of the circumferential-axial two-dimensional temperature field of the surface of slender objects. Attached Figure Description

[0015] Figure 1 A schematic diagram of a circumferential motion tracking and scanning infrared measurement device for measuring slender, high-temperature deformed pipes.

[0016] Figure 2 A schematic diagram showing the positions of sixteen thermopile array infrared sensors and two cameras;

[0017] Figure 3 Create a schematic diagram of the pixel coordinate system and the camera coordinate system;

[0018] Figure 4 This is a flowchart of the algorithm for locating the tested pipe in an image;

[0019] Figure 5 A schematic diagram illustrating the acquisition of the three-dimensional coordinates of the pipe under test using the pixel coordinates of two cameras;

[0020] Figure 6 A schematic diagram of the temperature data of the measured pipe acquired by the infrared sensor of the thermopile array;

[0021] Figure 7 This is a flowchart of the host computer's workflow. Detailed Implementation

[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. These examples are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0023] like Figure 1 As shown in the figure, this application provides an infrared measurement device for circumferential temperature field distribution of slender objects, including: a thermopile array infrared sensor 1, an industrial camera 2, a mounting bracket 3, a ring circuit board 4, a three-dimensional motion platform 5, an infrared control box 6, a switch 7, and a pipe under test 8.

[0024] In this embodiment, sixteen thermopile array infrared sensors 1 are connected to a ring circuit board 4 via signal lines, and are fixed to a mounting bracket 3 with bolts. Two industrial cameras 2 and the ring circuit board 4 are also fixed to the mounting bracket 3 with bolts. The mounting bracket 3 is fixed to a three-dimensional motion platform 5. The sixteen thermopile array infrared sensors 1 are connected to the ring circuit board 4 via data lines and transmit data information. The infrared control box 6 is connected to the three-dimensional motion platform 5 via data lines and transmits data information. The two industrial cameras 2, the ring circuit board 4, and the infrared control box 6 are connected to a host computer via data lines and transmit data information. The host computer controls the three-dimensional motion platform 5 to perform axial scanning (resolution 0.1 mm) and radial two-dimensional tracking of the measured pipe 8, ensuring that the sensor center is aligned with the axis of the measured pipe 8.

[0025] In some embodiments, sixteen thermopile array infrared sensors (8×8 pixels) are circumferentially distributed at equal angles, covering the 4.7~5.3 μm band, with a frame rate of 88 Hz, a temperature measurement range of -20~1000℃, and a field of view of 23°. The thermopile array infrared sensor can not only acquire the overall temperature distribution of the target area but also accurately locate local hotspots, improving the spatial resolution of temperature monitoring.

[0026] Furthermore, the output mode of the aforementioned sensor is infrared radiation temperature. This signal, after calibration, has high temperature measurement accuracy and repeatability, making it suitable for high-precision non-contact temperature measurement.

[0027] In some embodiments, the pipe under test is heated by a heating source that provides a stable current, enabling the heating system to gradually heat the pipe 8 under test according to a predetermined power, ensuring that the temperature field remains stable during the heating process. By precisely adjusting the output power and current of the heating source, the pipe 8 under test can be stabilized in different temperature fields, avoiding measurement errors introduced by temperature fluctuations.

[0028] In some embodiments, the industrial camera has a pixel resolution of 1280×1024 and a frame rate of 91.3 Hz, capturing more than 1.3 million pixels of detail per frame. These pixels accurately reproduce the details of the object being photographed. Through high-resolution image capture, the industrial camera can precisely record the surface features and minute changes (deformations) of the measured pipe, thereby achieving high-resolution image capture.

[0029] This application also provides an infrared measurement method for the circumferential temperature field distribution of slender objects, including:

[0030] 1. Dynamic tracking and positioning

[0031] A mapping relationship between the pixel coordinate system and the three-dimensional spatial coordinate system of the pipeline is established through dual-camera calibration. After calibration, a real-time image processing algorithm is used to acquire the edge center position of the measured pipeline from the dual cameras. Utilizing the principle of binocular vision, the three-dimensional coordinates of the pipeline edge are calculated by matching feature points in the image. Based on these coordinate information, the midpoint of the common perpendicular of the pipeline is calculated to determine the three-dimensional spatial position of the pipeline. Real-time deformation monitoring is performed in conjunction with image processing feedback to calculate the pipeline deformation offset and drive the motion platform for dynamic compensation, ensuring the accuracy of the measurement results.

[0032] 2. Temperature field reconstruction

[0033] Sixteen thermopile array infrared sensors synchronously acquire circumferential temperature data of the pipe surface. The accuracy of the data from each sensor is ensured through radiation source size effect correction and consistency calibration. Then, the data from multiple sensors are fused based on a deconvolution algorithm to reconstruct the continuous circumferential temperature distribution of the pipe surface, and a complete two-dimensional temperature field is generated by combining it with axial scan data.

[0034] Based on the above basic concept, the specific steps of the embodiments of this application are as follows:

[0035] Step 1: Fix sixteen thermopile array infrared sensors 1, two industrial cameras 2 and a ring circuit board 4 on the mounting bracket 3, and fix the bracket 3 on the three-dimensional motion platform 5.

[0036] The thermopile array infrared sensors 1 are distributed at equal angular intervals around the mounting bracket 3, with the probes of the thermopile array infrared sensors 1 100 mm from the center of the mounting bracket 3. Two industrial cameras 2 are orthogonally and symmetrically distributed, with equal distances from the center of the mounting bracket 3 to both industrial cameras 2. The installation layout of the sixteen thermopile array infrared sensors 1 and the two industrial cameras 2 is as follows: Figure 2 As shown, the angle between the two industrial cameras 2 is 90°, and the sixteen thermopile array infrared sensors 1 are arranged at equal intervals around the object being measured, with the angle between each pair being 22.5°.

[0037] The thermopile array infrared sensor 1 measures the temperature of an object by detecting the infrared radiation emitted by the object. When the sensor receives infrared radiation from objects at different temperatures, the thermocouples generate a voltage. This voltage is proportional to the temperature, and the object's temperature can be obtained by calculating the magnitude of the voltage. Multiple thermocouples arranged in an array can simultaneously detect the temperature at different locations, thus forming a temperature distribution map.

[0038] The two industrial cameras 2 primarily form images by capturing light reflected from objects. They focus light through lenses onto an image sensor, which converts the light signals into electronic signals, which are then processed into a digital image. Figure 2 The two industrial cameras are configured at a 90° angle to achieve a wider field of view and more accurate stereo vision measurements.

[0039] Specifically, the two industrial cameras 2 are placed perpendicular to each other, allowing the system to acquire image data of the object from two different perspectives. This is very helpful for 3D reconstruction, depth measurement, and accurate spatial positioning of the object. The sixteen thermopile array infrared sensors 1 are arranged at a 22.5° angle between each pair, typically to ensure that the sensors can cover a wider area and obtain more uniform temperature measurement data. The ring circuit board 4 is fixed to the mounting bracket 3 with bolts. A host computer controls the movement of the 3D motion platform 5 to center the measured pipe 8 on the mounting bracket 3.

[0040] Step 2: Establish the two-dimensional pixel coordinate system and the three-dimensional camera coordinate system of industrial camera 2.

[0041] The structure of the corresponding coordinate system is as follows Figure 3 As shown, 3a represents the two-dimensional pixel coordinate system of industrial camera 2, and 3b represents the three-dimensional camera coordinate system of industrial camera 2; where the two-dimensional pixel coordinate system... The image acquired by industrial camera 2 is represented by two-dimensional coordinates, with the top-left corner of the image as the origin. The three-dimensional camera coordinate system... It is a three-dimensional coordinate system with the center of the optical system of industrial camera 2 as the origin, where The axis is the camera's optical axis, where axis The axes correspond to the U-axis and V-axis of the pixel coordinate system, respectively. The transformation formula between homogeneous coordinates in the two-dimensional pixel coordinate system and homogeneous coordinates in the three-dimensional camera coordinate system is shown in Equation 1.

[0042]

[0043] in This is the intrinsic parameter matrix of the camera. , The pixel is the unit of focal length for a camera. , This refers to the pixel position at the center of the image, that is, the pixel corresponding to the optical axis.

[0044] Step 3: Use the calibration board to perform single-target and dual-target calibration on the two industrial cameras 2.

[0045] Two industrial cameras 2 were used to acquire clear images of the calibration plate from different angles. Single-target positioning processing software was used to calculate the corresponding intrinsic parameter matrices of the two industrial cameras 2. The intrinsic parameter matrices describe the internal parameters of each camera.

[0046] Then, the calibration plate is fixed so that both industrial cameras 2 can simultaneously acquire clear images of the calibration plate. Using the 3D coordinates of the calibration plate, the transformation matrix from the two 3D camera coordinate systems to the calibration plate coordinate system is calculated using a calibration algorithm. Then, using these two transformation matrices, the relative transformation matrix between the two camera coordinate systems can be calculated. Where R is the rotation matrix and T is the offset matrix.

[0047] Step 4: The two-dimensional pixel coordinates of the pipe 8 under test, acquired and established by the two industrial cameras 2, are converted into the three-dimensional coordinates of the pipe 8 under test, as shown in the figure. Figure 5 As shown. Two industrial cameras 2 are positioned at a 90° angle in space to capture images of the pipe 8 under test, which can more accurately obtain the three-dimensional information of the pipe 8 under test. The captured images represent the planar position of the pipe 8 under test through two-dimensional pixel coordinates. The three-dimensional spatial position of the pipe 8 under test can be calculated by using the three-dimensional camera coordinate system established in step 2 and the camera intrinsic parameter matrix obtained by camera calibration in step 3.

[0048] Furthermore, the flowchart of the algorithm for identifying and locating the position of the tested pipe 8 in the image is as follows: Figure 4 As shown.

[0049] 1. Preprocess the image by using a Gaussian template to perform low-pass filtering to remove noise from the image and analyze the pixel histogram of the image.

[0050] 2. Because the tested pipe 8 is darker than the background in the image, and the number of pixels of the tested pipe 8 in the image is often fixed, an adaptive threshold is calculated based on the histogram, and the image is binarized. Pixels with values ​​below the threshold are recorded as 1, and pixels with values ​​above the threshold are recorded as 0. Then, the image is divided into hundreds of small regions, and the number of pixels with values ​​of 1 in each small region is counted. Regions where the proportion of 1s exceeds one-third are recorded as pixel regions.

[0051] 3. Perform morphological processing on the pixel region to expand the region and smooth the boundary; then cut out the region of the pipe under test 8 from the whole image to reduce the amount of computation in subsequent positioning calculations; use an edge detection algorithm to detect the upper and lower edges of the pipe under test 8, and average the positions of the upper and lower edges of the pipe under test 8 to obtain the center position of the pipe under test 8.

[0052] 4. After obtaining the preliminary center position of the pipe 8 under test, the intrinsic parameter matrix of the camera can be obtained based on single-target camera calibration. The set of position points of the pipe 8 under test in the images acquired by the two industrial cameras 2 is mapped to the ray sets of the two 3D camera coordinate systems. Then, the transformation matrix of the two 3D camera coordinate systems is obtained through dual-target calibration, and the ray set in one 3D camera coordinate system is mapped to the ray set in the other 3D camera coordinate system. The distances from all rays in each ray set to all rays in the other ray set are calculated. The ray pair with the shortest distance between rays is selected, and the midpoint of the common perpendicular of the ray pair is calculated. The midpoint of the common perpendicular is defined as the center position of the pipe 8 under test. The center position of the pipe 8 under test is then precisely corrected.

[0053] Step 5: Scan the pipe 8 under test, correct the data obtained from the scan, and restore the circumferential temperature field of the pipe 8 under test. Figure 6 This is a schematic diagram of temperature data acquired by a thermopile infrared sensor from the tested pipe 8. The horizontal axis represents the position of a single thermocouple. The sixteen thermocouples are arranged in a ring, with an angular interval of 22.5° between each thermocouple. Because the data is mirrored and repeated when the pipe is symmetrical, the data is not identical to the other half. Figure 6 Only half a circle of thermocouples (8 in total) are shown to save memory space. Figure 6 The vertical axis represents 8 consecutive sampling time points, with a time interval of 50ms. The data in each cell represents the temperature value collected by the thermocouple, and the shade of color represents the temperature level, with darker colors indicating higher temperatures and lighter colors indicating lower temperatures.

[0054] The flowchart of scanning the tested pipeline 8 using the host computer control device is as follows: Figure 7As shown in the diagram, the host computer is first used to connect the ring circuit board 4, the infrared control box 6, and the two industrial cameras 2. After successful connection, data acquisition begins. This process includes acquiring data from sixteen thermopile array infrared sensors 1, image data from the two industrial cameras 2, and pulse position data from the three axes of the infrared control box 6. The three-dimensional coordinate position of the measured pipe 8 in the camera coordinate system is calculated using the images acquired by the two industrial cameras 2.

[0055] 1. Based on the three-dimensional coordinates of the tested pipe 8 in the camera coordinate system, the host computer controls the infrared control box 6 to output pulse signals to the three-dimensional motion platform 5. The three-dimensional motion platform 5 moves to keep the tested pipe 8 at the center of the mounting bracket 3. After the scan length is input in the host computer, the two industrial cameras 2 begin to track the pipe. After the target tracking stabilizes, the sixteen thermopile array infrared sensors 1 begin to collect and save the circumferential temperature data of the tested pipe 10. After the data is saved, a fixed-length scan is performed. After the scan is completed, the experimental platform returns to its original position to perform the next scan or stop scanning.

[0056] 2. After the scan is completed, the position of the tested pipe 8 is calculated based on the data recorded by the host computer, and the temperature data measured by the thermopile array infrared sensor 1 is obtained. Radiation source size effect correction is performed based on the relative position of the tested pipe 8 and the thermopile array infrared sensor 1 (the correction parameters need to be calibrated offline in advance). Next, temperature measurement consistency correction is performed by the thermopile array infrared sensor 1, and finally the radiation temperature of the sixteen surfaces of the outer wall of the tested pipe 8 is obtained.

[0057] 3. Finally, the radiation temperature data of the sixteen surfaces are first fused using a linear interpolation formula, as shown in Equation 2:

[0058]

[0059] In this formula and These are the coordinates of two adjacent thermocouples. and These are the temperature values ​​collected by two thermocouples. It represents a certain angle value within the angle between two thermocouples. This indicates the temperature value corresponding to that angle. This is used to calculate the temperature measurement value around the entire perimeter of the tested pipe.

[0060] The actual circumferential temperature distribution of the outer wall of the tested pipe 8 is then calculated using deconvolution. The deconvolution formula is shown in Equation 3:

[0061]

[0062] This represents the actual temperature distribution in the circumferential direction. This represents the temperature value collected by the thermocouple. This represents the convolution kernel. The deconvolution operation is used to recover the original temperature signal that is closest to the true value.

[0063] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. An infrared measurement method for the circumferential temperature field distribution of a slender object, wherein the circumferential temperature distribution of the slender object is non-uniform, characterized in that, Includes the following steps: Dynamic tracking and positioning: At least two vision sensors are used to acquire image information of the slender object being measured. The three-dimensional spatial position and deformation information of the slender object being measured are determined by the image processing algorithm. The motion platform is driven to make dynamic adjustments based on the position and deformation information so that the measuring device and the slender object being measured maintain a relatively stable positional relationship. Ensure that the center of the multiple circumferentially distributed sensors is aligned with the axis of the slender object being measured; Temperature data acquisition: Multiple infrared sensors are distributed around the circumference of the slender object being measured to simultaneously collect temperature data of its surface. Data processing and temperature field reconstruction: The collected temperature data is corrected and then the data from multiple infrared sensors are fused through a data fusion algorithm to reconstruct the circumferential continuous temperature distribution on the surface of the slender object being measured, and a complete two-dimensional temperature field distribution is generated by combining the axial scan data. The data fusion algorithm uses a deconvolution algorithm, which initially fuses data from multiple infrared sensors using a linear interpolation formula, and then performs deconvolution calculation to obtain the actual circumferential temperature distribution of the outer wall of the slender object being measured. The image processing algorithm includes the following sub-steps: a. Preprocess the acquired images by using a low-pass filter to remove noise; b. Calculate the adaptive threshold based on the image pixel histogram and perform binarization; c. Perform region segmentation and morphological processing on the binarized image to extract the edge features of the slender object being measured; d. Determine the center position of the slender object being measured using an edge detection algorithm, and calculate its three-dimensional coordinates using the principle of binocular vision; It also includes real-time deformation monitoring of the slender object being measured, calculating the offset of the deformation, and driving the motion platform to perform dynamic compensation to ensure the accuracy of the measurement results.

2. The infrared measurement method for circumferential temperature field distribution according to claim 1, characterized in that, The vision sensor is an industrial camera, and the pixel resolution of the industrial camera is not less than 1280×1024, the frame rate is not less than 90 Hz, and the three-dimensional coordinates of the slender object being measured are calculated through the principle of binocular vision.

3. The infrared measurement method for circumferential temperature field distribution according to claim 1, characterized in that, The correction process includes radiation source size effect correction and sensor temperature measurement consistency correction.

4. The infrared measurement method for circumferential temperature field distribution according to claim 1, characterized in that, The number of infrared sensors is no less than 16, and they are evenly distributed around the circumference of the slender object being measured.

5. The infrared measurement method for circumferential temperature field distribution according to claim 4, characterized in that, The infrared sensor is a thermopile array infrared sensor with a pixel count of no less than 8×8, a coverage band of 4.7 to 5.3 μm, a frame rate of no less than 80 Hz, and a temperature measurement range of -20 to 1000℃.

6. The infrared measurement method for circumferential temperature field distribution according to claim 2, characterized in that, The number of vision sensors is two, and they are orthogonally symmetrically distributed at a 90° angle.

7. The infrared measurement method for circumferential temperature field distribution according to claim 1, characterized in that, It also includes heating the slender object being tested to create a stable temperature field on its surface, the heating being achieved by providing a stable current through a heating source.

Images