Binocular vision point cloud temperature field rapid reconstruction method based on homography transformation

Abstract

The invention provides a binocular vision point cloud temperature field rapid reconstruction method based on homography transformation. During calibration, the binocular camera and the thermal infrared imager shoot mark points synchronously; solving a homography matrix between the left camera and the thermal infrared imager according to a mark point visible light image obtained by the left camera and a corresponding point in the thermal image; and based on the homography matrix, resolving an external parameter matrix between the left camera and the thermal infrared imager. When a point cloud temperature field is reconstructed, a binocular camera obtains a surface speckle image of a measured object and calculates a three-dimensional point cloud on the surface of the measured object; updating a homography matrix by using plane features in the three-dimensional point cloud; and according to the internal reference matrix and the updated homography matrix, mapping the temperature information in the thermogram to a point cloud corresponding to the visible light coordinates of the left camera. According to the method, only one-time calibration measurement is needed, operation is easy, precision is high, and measurement of the three-dimensional morphology and the temperature of each point of the material with unobvious surface features is achieved.

Description

Technical Field

[0001] The invention belongs to the field of computer graphics, and in particular relates to a method for quickly reconstructing a binocular vision point cloud temperature field based on homography transformation. Background Art

[0002] In the current industrial and scientific research fields, accurate measurement and analysis of the three-dimensional morphology and temperature field of high-temperature objects has become a key requirement. In industrial inspection, material science, and physical research in high-temperature environments, it is crucial to accurately measure and analyze the temperature distribution on the surface of an object. Although traditional temperature measurement technologies, such as thermocouples and infrared thermometers, can meet the needs of temperature detection to a certain extent, they have certain limitations in terms of complex surfaces, non-contact measurement, and high-resolution temperature field acquisition. Especially in high-temperature environments, due to the enhancement of thermal radiation and the influence of complex surface geometry, traditional temperature measurement methods often find it difficult to provide sufficient accuracy and spatial resolution.

[0003] In recent years, with the development of computer vision technology, binocular vision-based measurement methods have gradually become one of the important means of high-precision, non-contact temperature field measurement. The binocular vision system uses two cameras to capture images of the same scene from different angles and reconstructs the three-dimensional shape of the object through the principle of triangulation. Combined with image processing technology, this method can accurately obtain three-dimensional point cloud data on the surface of an object in a high-temperature environment, providing a geometric basis for further temperature field analysis.

[0004] However, in practical applications, relying solely on the reconstruction and fusion of geometric morphology cannot fully reflect the temperature distribution on the surface of an object. Especially in high-temperature environments, a small change in the surface temperature of an object may lead to a large thermal effect. Therefore, how to effectively combine the temperature field with three-dimensional point cloud data has become a key issue in current research. Traditional temperature field reconstruction methods, such as thermal field analysis based on infrared images, often rely too much on the geometric features of the object surface, resulting in low accuracy in temperature field reconstruction, which is difficult to meet the needs of measuring flat test blocks and smooth surfaces.

[0005] Three-dimensional digital image correlation (3D-DIC) based on binocular vision can analyze the speckle pattern on the surface of an object and reconstruct the point cloud, but it cannot obtain temperature information.

[0006] The patent application with publication number CN110379002A, "A method for displaying surface temperature in three-dimensional reconstruction based on infrared and visible light image fusion", first calculates the scale factor of the space object between the infrared and visible light cameras, then uses the natural color transmitted by the YUV space color to fuse the target image, and finally replaces the original visible light image in the process of reconstructing the dense point cloud of the three-dimensional reconstruction to reconstruct the surface, so as to realize the three-dimensional reconstruction of the surface with temperature information. The scale factor and relative offset extracted by this fusion method are only applicable to the calibration plane, and no factor correction is performed for the plane pose when directly brought into the measurement plane. When fusing the infrared camera image and the visible light image, pixel errors will be introduced due to the difference in offset and resolution between the two. Using the fused image for dense point cloud reconstruction will bring this part of the error into the point cloud coordinates, affecting the coordinate accuracy.

[0007] The patent application with publication number CN110379002A, “A coupled measurement method and system for strain field and temperature field that fuses infrared information”, adopts the following means to fuse the strain field and temperature field: the infrared image is processed by bilinear interpolation to compensate for the resolution difference between the three-dimensional strain map and the infrared temperature field map of the speckle sample at different times, so that the two images are the same size. The temperature and strain at any position of the material can correspond one by one, so as to obtain the corresponding three-dimensional strain field and temperature field at each time, and realize the coupling of the strain field and the temperature field. The measurement system only shows that the three cameras are calibrated using a dot calibration plate containing three concentric circles, but fails to point out the calibration steps and algorithms, which are the key to the precise coupling of the strain field and the temperature field. In addition, the fusion of the strain field and the temperature field occurs in the final stage, and the original measurement image is not fully utilized. At the same time, the system needs to be calibrated before each measurement, and the process is cumbersome and complicated.

[0008] In order to meet these challenges, it is of great significance to fuse point cloud and temperature field and simplify the calibration and solution process. Summary of the invention

[0009] In view of this, the present invention provides a method for rapid reconstruction of temperature field of scattered spot cloud based on homography transformation. A system is formed by an infrared thermal imager and a visible light binocular vision camera to measure the surface speckle image of the object at the same time, and the relationship between visible light and thermal image is mapped by preliminary calibration calculation to realize the fusion of point cloud and temperature. The proposed method only requires a single calibration measurement, is simple to operate, has high accuracy, and can realize the measurement of the three-dimensional morphology and temperature of each point of materials with unclear surface features.

[0010] In order to solve the above technical problems, the present invention is implemented as follows.

[0011] A method for rapid reconstruction of binocular visual point cloud temperature field based on homography transformation, the method is based on a measurement system consisting of a group of binocular cameras and an infrared thermal imager adjacent to the left camera of the binocular cameras; the measurement system obtains the visible light image of the photographed object through the binocular camera and solves the three-dimensional point cloud; the temperature distribution information of the object surface is captured by the infrared thermal imager; the method includes:

[0012] Step 1: Calibration of the left camera and the infrared thermal imager: The binocular camera and the infrared thermal imager synchronously shoot the marker points; solve the homography matrix H between the left camera and the infrared thermal imager according to the corresponding points in the visible light image and the thermal image of the marker points obtained by the left camera; obtain the plane features according to the three-dimensional point cloud solved by the binocular camera image; based on the homography matrix H and the plane features, combined with the intrinsic parameter matrices of the left camera and the infrared thermal imager, solve the extrinsic parameter matrix between the left camera and the infrared thermal imager;

[0013] Step 2: Point cloud temperature field reconstruction: The binocular camera obtains the surface speckle image of the measured object and solves the three-dimensional point cloud of the measured object surface; the plane features in the three-dimensional point cloud are used to update the homography matrix H; according to the intrinsic parameter matrix and the updated homography matrix H, the temperature information in the thermal image is mapped to the point cloud corresponding to the visible light coordinates of the left camera, completing the fusion of the temperature field and the three-dimensional point cloud.

[0014] Preferably, in step 1, the homography matrix H between the left camera and the infrared thermal imager is solved according to the corresponding points in the visible light image of the marker points obtained by the left camera and the thermal image: the homography matrix H is obtained by fitting the plane using the least squares method.

[0015] Preferably, in step 1, based on the homography matrix H and the plane features, combined with the intrinsic parameter matrices of the left camera and the infrared thermal imager, the extrinsic parameter matrix between the left camera and the infrared thermal imager is solved as follows:

[0016] The coordinates of a point in space P = [X, Y, Z] T , the pixel coordinates in the left camera image are [u 1 ,v 1 ], the pixel coordinates in the thermal image are [u 0 ,v 0 ]; written as homogeneous coordinates represented by p 1 =[u 1 ,v 1 ,1]T,p 0 =[u 0 ,v 0 ,1]T;

[0017] According to binocular stereo vision, the relationship between pixel coordinates and world coordinates is as follows:

[0018] p 1 =K1 P,p 0 =K 0 (RP+t)

[0019] Among them, K 1 and K 0 are the intrinsic parameter matrices of the left camera and the infrared thermal imager respectively; R and t are the rotation matrix and translation vector in the extrinsic parameter matrix between the left camera and the infrared thermal imager respectively;

[0020] In the world coordinate system, the target plane equation is expressed as:

[0021] n T Pd=0

[0022] Point p 1 to p 0 The homography transformation is described as p 0 =Hp 1 , combined with p 0 and p 1 The expression of and the target plane equation, the expression of the homography transformation matrix H is as follows:

[0023]

[0024] The 3D point cloud obtained by the binocular camera is used to calculate the calibration plane normal vector n = [n x ,n y ,n z ] T Substitute the distance d from the plane to the origin of the coordinate system into the H expression (1) and solve to obtain the transformation matrix R and t from the infrared thermal imager to the left camera.

[0025] Preferably, in step 2, the temperature information in the thermal image is mapped to the point cloud corresponding to the visible light coordinates according to the rotation matrix R, the translation vector t and the updated homography matrix H, as follows:

[0026] Thermal image coordinates p 0 and the visible light coordinates p of the left camera 1 The mapping relationship is:

[0027]

[0028] According to the above formula, the temperature information in the thermal image is first mapped to the visible light coordinates; then based on the corresponding relationship between the left camera coordinates and the three-dimensional point cloud coordinates, the mapping of the temperature information to the three-dimensional point cloud is realized.

[0029] Preferably, in step 2, when mapping is performed, interpolation calculation is performed on the temperature information of the thermal image, and the temperature information in the thermal image is accurately mapped to the point cloud corresponding to the visible light coordinates of the left camera.

[0030] Preferably, the interpolation calculation adopts a bicubic difference algorithm.

[0031] Preferably, in step 2, before mapping the temperature information of the thermal image, the thermal image is further subjected to distortion correction, resolution adjustment and / or noise filtering.

[0032] Preferably, in step 2, the binocular camera acquires a surface speckle image of the measurement object, and solves the three-dimensional point cloud of the measurement object surface by analyzing the displacement change of the speckle through a three-dimensional digital image correlation 3D-DIC method to solve the three-dimensional point cloud of the measurement object surface.

[0033] Preferably, the optical axes of the left camera and the infrared thermal imager are parallel.

[0034] Beneficial effects:

[0035] (1) The present invention effectively combines the temperature field information captured by the infrared thermal imager with the high-precision three-dimensional point cloud data obtained by the binocular vision system through homography transformation. Compared with the existing technology, it can achieve high-precision combination of temperature and morphology after single-frame calibration, reduce the measurement error caused by the resolution difference between different devices, and thus improve the accuracy of temperature field reconstruction.

[0036] (2) When calibrating the extrinsic matrix, the present invention only needs to fix the measurement system camera, use a single image to obtain the homography matrix, and combine it with the intrinsic matrix to obtain the calibration result of the extrinsic matrix, without using multiple images to calculate the extrinsic matrix. This simplifies the operation steps, reduces the burden on users, and improves the practicality of the system while ensuring accuracy.

[0037] (3) The present invention is applicable to the measurement of the surface temperature field of objects under different temperature environments. It can adapt to the measurement of the surface of objects without obvious features, and does not require the feature matching of the visible light image and the thermal image. At the same time, the bicubic interpolation algorithm is used to accurately map the temperature information of the thermal image to the point cloud data, further reducing the error caused by resolution mismatch, ensuring the fusion of the temperature field and the three-dimensional morphology at any scale, and providing a more accurate temperature field reconstruction result. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is a flow chart of the method for reconstructing the three-dimensional point cloud and temperature field of the object surface constructed by the present invention;

[0039] Figure 2 It is a specific implementation flow chart of the calibration of the object surface three-dimensional point cloud and temperature field reconstruction measurement system constructed by the present invention;

[0040] Figure 3 It is a schematic diagram of the object surface three-dimensional point cloud and temperature field reconstruction measurement system constructed by the present invention;

[0041] Figure 4 It is a specific example of camera measurement and acquisition. Figure (a) is an image taken by a visible light camera, and Figure (b) is a 3D point cloud and temperature color scale diagram of the reconstructed surface with temperature information.

[0042] Among them: 1-controller, 2-left camera in the binocular camera, 3-right camera in the binocular camera, 4-infrared thermal imager, 5-tripod, 6-test furnace. DETAILED DESCRIPTION

[0043] In order to make the purpose, technical solution and advantages of the present invention more clear, the present invention is further described in detail in conjunction with the accompanying drawings and embodiments. The present invention provides a binocular vision scattered spot cloud temperature field reconstruction method based on homography transformation, which combines the measurement data of the binocular camera and the infrared thermal imager, and realizes the efficient fusion of the point cloud and the temperature field based on homography transformation, which is suitable for the synchronous measurement and mapping of the surface temperature and three-dimensional morphology of an object.

[0044] Figure 3 FIG. 2 shows the measurement system on which the reconstruction method of the present invention is based. Figure 3 As shown, the measurement system includes a binocular camera, an infrared thermal imager 4 and a controller 1. The binocular camera includes a left camera 2 and a right camera 3. The infrared thermal imager 4 is adjacent to the left camera 2. In this embodiment, for ease of installation, the binocular camera and the infrared thermal imager 4 are both installed on the same tripod 5. The controller 1 is connected to the binocular camera and the infrared thermal imager 4 to obtain the collected image for processing. The measurement object is arranged in the test furnace 6.

[0045] When the measurement system collects images, it uses a binocular camera and an infrared thermal imager to synchronously shoot the target object. The binocular camera is responsible for obtaining the three-dimensional shape information of the object, while the infrared thermal imager captures the temperature distribution on the surface of the object.

[0046] In order to ensure the accuracy and synchronization of the data, attention should be paid to the configuration and installation of the camera group, as well as the synchronous control of image acquisition. Specifically, the left camera, the right camera and the infrared thermal imager are fixed on the same measuring platform, and the distance between the left camera and the infrared thermal imager is close to reduce the parallax between the two. The positional relationship between the binocular camera and the infrared thermal imager is fixed to ensure the geometric correspondence between the two. However, the spatial positional relationship between the two is not limited. It is advisable that the optical axes of the left camera and the thermal imager are parallel. In the case of installation restrictions, they can be approximately parallel. The specific position can be achieved through precise brackets and adjustment mechanisms to improve the consistency of image acquisition. The schematic diagram of the object surface three-dimensional point cloud and temperature field reconstruction measurement system is shown in the figure. Figure 3 shown.

[0047] For synchronous control of image acquisition, use hardware synchronization triggers or control software to trigger the binocular camera and infrared thermal imager at the same time to ensure that they acquire images at the same time point.

[0048] In this embodiment, if Figure 1 As shown, the present invention is directed to a method for reconstructing a three-dimensional point cloud and temperature field of an object surface based on homography transformation, comprising the following steps:

[0049] Step S1: System calibration

[0050] The calibration process of the object surface three-dimensional point cloud and temperature field reconstruction solution based on homography transformation of the present invention is as follows: Figure 2 As shown in Figure 2, system calibration is a key step to achieve accurate mapping between 3D point cloud and temperature field. By jointly calibrating the binocular camera and the infrared thermal imager, the homography transformation matrix between the two can be obtained.

[0051] The calibration process specifically includes the following steps:

[0052] Step S1.0: Arrangement of marker points: Use a standard plane as a calibration plate, and arrange marker points with high reflectivity on the surface of the standard plane, such as reflective patches or infrared LED dot matrix. These marker points will be used for image registration and calibration in subsequent steps. The selected marker points should be clearly recognizable in both visible light and infrared images under natural light or external infrared light source. The reflective marker points should be evenly distributed on the surface of the standard plane, and the number should be sufficient to meet the needs of subsequent homography matrix calculation. In this embodiment, the number of marker points should be no less than 4, and the spatial position of the marker points should be disordered, asymmetric, and non-collinear. The position of the marker points in the image will be used to calculate the external parameter matrix of the camera.

[0053] Step S1.1: Calibration of the binocular camera and thermal imager: This is to calibrate the intrinsic parameter matrix. The intrinsic parameter matrix K of the binocular camera 1 The calibration is done by the classic Zhang Zhengyou calibration method or other calibration techniques based on checkerboard images, and the intrinsic parameter matrix K of the thermal imager is 0 It is accurately determined through a special thermal imaging calibration method.

[0054] Step S1.2: The binocular camera and the infrared thermal imager respectively take images of the calibration plate: the binocular camera and the infrared thermal imager are triggered synchronously, the binocular camera collects visible light images, and the three-dimensional point cloud is obtained by registration and solution; the infrared thermal imager collects thermal images; in order to ensure the accuracy of the calibration, it should be ensured that each camera and thermal imager can see at least four or more marking points.

[0055] Step S1.3: Solve the homography matrix:

[0056] The coordinates of a point in space P = [X, Y, Z]T , the pixel coordinates in the left camera image are [u 1 ,v 1 ], the pixel coordinates in the thermal image are [u 0 ,v 0 ] written as homogeneous coordinates represented by p 1 =[u 1 ,v 1 ,1]T,p 0 =[u 0 ,v 0 ,1]T.

[0057] Point p 1 to p 0 The homography transformation can be described as follows.

[0058] p 0 =Hp 1

[0059] Where H is the homography matrix between the left camera and the infrared thermal imager;

[0060] In the actual calibration process, there is good correspondence between the marker points. Therefore, the present invention uses a simple least square method to construct a set of equations to linearly estimate the homography matrix H. The least square method is used to fit the plane to obtain the H matrix, which can reduce the calibration error and improve the accuracy of the homography transformation.

[0061] Will

[0062]

[0063] Since point p 0 and p 1 Using homogeneous coordinates, the homography matrix has only 8 degrees of freedom, and one pair of matching points can construct two constraints, so at least four pairs of matching points are required to solve the homography matrix.

[0064] Substitute the homography transformation expression and sort it out to get a pair of matching point equations:

[0065]

[0066] nThe above matching point equations are combined and recorded as:

[0067] Ah=0

[0068] Among them, the size of A is 2n×9, and h is the column vector corresponding to H, with a size of 9×1. Applying singular value decomposition, we get

[0069] A=U∑V T

[0070] Among them, U, V, and Σ are two orthogonal matrices and a diagonal matrix obtained by singular value decomposition. h corresponds to the last column of V, that is, the right singular vector corresponding to the minimum singular value. Thus, H can be obtained.

[0071] Step S1.4: plane feature extraction: According to the common point of the left and right cameras of binocular stereo vision, the three-dimensional coordinates of the calibration mark point can be obtained, and the normal vector n of the calibration plane can be determined. x ,n y ,n z ] T The distance d from the origin is the plane feature.

[0072] In three-dimensional space, a plane can be represented as:

[0073] n x x+n y y+n z zd=0

[0074] In the above formula, x, y, and z are the three-dimensional coordinates of the point cloud. In this step, the point cloud data of the three landmark points are substituted into the above formula, and the least square method is used to solve the plane fitting, and finally the plane normal vector n and the distance d from the plane to the origin of the world coordinate system are obtained.

[0075] Step S1.5: Calculate the external parameter matrices of the left camera and the thermal imager: R and t.

[0076] According to binocular stereo vision, the relationship between pixel coordinates and world coordinates is as follows:

[0077] p 1 =K 1 P,p 0 =K 0 (RP+t)

[0078] Among them, R and t are the external parameter matrices of the thermal imager relative to the left camera, R is the rotation matrix, and t is the translation vector. K 1 and K 0 are the intrinsic parameter matrices of the calibrated left camera and thermal imager respectively. In the world coordinate system, the target plane equation can be expressed as:

[0079] n T Pd=0

[0080] The p 0 and p 1 Substituting the expression of into the equation, we can get the expression of the homography transformation matrix as follows:

[0081]

[0082] Using the calibration plane normal vector n = [n x ,ny ,n z ] T Substituting the distance d from the plane to the origin of the coordinate system into the H expression, we can obtain the transformation matrix R and t from the thermal imager to the left camera.

[0083] Step S2: Point cloud reconstruction and temperature field mapping.

[0084] Step S2.1: Capture an image. In this step, a binocular camera is used to obtain a visible light image of the surface speckle of the measured object, and the three-dimensional point cloud morphology of the surface of the measured object is solved. In practice, a speckle texture can be prepared on the surface of the measured object, and a binocular camera can be used to capture speckle images from different angles to generate a three-dimensional point cloud; or the ADIC3D library of Matlab can be used to analyze the displacement change of the speckle through the three-dimensional digital image correlation 3D-DIC method, solve the visible light image of the speckle, and generate a three-dimensional point cloud.

[0085] Step S2.2: Update the homography matrix H to map the temperature information in the thermal image to the point cloud corresponding to the visible light coordinates of the left camera.

[0086] The planes of the calibration process and the measurement process are different, resulting in different homography matrices H. Therefore, it is necessary to recalculate the plane features and update the homography matrix H.

[0087] The calculation formula of the plane feature is the same as step 1.4, except that the three-dimensional coordinates of multiple points in the three-dimensional point cloud information obtained in step S2.1 are used for fitting. Due to the influence of the rough surface of the workpiece and the noise of the solution, the RANSAC algorithm can be further used to eliminate the noise and improve the plane fitting accuracy. Then the new plane feature is used to update the homography matrix H.

[0088] Since the left camera and the infrared thermal imager actually constitute another binocular vision system, the transformation matrix R and t from the thermal imager to the left camera have been obtained in the calibration. After the homography transformation of the thermal imager pixels with the left camera pixels is completed, they can be unified into the binocular vision system. The point cloud and thermal imager pixel matching process is completed through point cloud-left camera pixel mapping and left camera-thermal imager pixel matching. Substituting the expression of H into the homography transformation expression, we can get:

[0089]

[0090] Thus, the temperature information in the thermal image can be mapped to the visible light coordinates. Due to the correspondence between the left camera coordinates and the three-dimensional point cloud coordinates, the mapping of temperature information to the three-dimensional point cloud can be achieved.

[0091] Since the resolution of thermal images and point clouds do not match, direct mapping may lead to large errors. The test block area measured in the 3D reconstruction of the point cloud temperature field is usually small, with a smooth surface and fewer features, and the temperature changes are relatively gentle. In order to reduce the complexity of calculations and improve the efficiency of the solution, the temperature information in the thermal image can be interpolated and then accurately mapped to each point cloud point to complete the fusion of the temperature field and the 3D point cloud. The interpolation algorithm can use the bicubic interpolation algorithm.

[0092] To further reduce the computational complexity, the thermal image may be preprocessed before mapping, such as adjusting the resolution and filtering the noise, to improve the efficiency of the bicubic interpolation calculation.

[0093] p 0 The temperature at is calculated as follows:

[0094]

[0095] Among them, (i,j) is the distance from (u 0 ,v 0 ) The nearest thermal image pixel coordinates, W is the kernel function, here we take

[0096]

[0097] Bicubic interpolation achieves a good balance between accuracy, smoothness and computational efficiency. The mapping process of temperature data in the present invention ensures good real-time performance.

[0098] The thermal images collected by the infrared thermal imager are spatially interpolated using the bicubic interpolation method, which overcomes the resolution difference and decimal coordinates between the infrared thermal imager and the two visible light cameras, so that the point cloud points can complete the accurate mapping of the temperature. The three-dimensional visualization screenshot of the point cloud temperature field after interpolation mapping is shown in the figure. Figure 4 As shown, the temperature at any position of the object can correspond to the temperature one by one, realizing the rapid reconstruction of the point cloud temperature field.

[0099] The above specific embodiments only describe the design principle of the present invention. The shapes and names of the components in the description may be different and are not limited. Therefore, those skilled in the art in the field of the present invention may modify or replace the technical solutions recorded in the above embodiments; and these modifications and replacements do not deviate from the creative purpose and technical solutions of the present invention and should all fall within the protection scope of the present invention.

Claims

1. A method for fast reconstruction of binocular visual point cloud temperature field based on homography transformation, characterized in that: The measurement system based on this method consists of a set of binocular cameras and an infrared thermal imager adjacent to the left camera of the binocular cameras; the measurement system obtains the visible light image of the object through the binocular cameras and solves the three-dimensional point cloud; Capturing temperature distribution information on the surface of an object by using an infrared thermal imager; the method comprises: Step 1: Calibration of the left camera and the infrared thermal imager: The binocular camera and the infrared thermal imager synchronously shoot the marker points; solve the homography matrix H between the left camera and the infrared thermal imager according to the corresponding points in the visible light image and the thermal image of the marker points obtained by the left camera; obtain the plane features according to the three-dimensional point cloud solved by the binocular camera image; based on the homography matrix H and the plane features, combined with the intrinsic parameter matrices of the left camera and the infrared thermal imager, solve the extrinsic parameter matrix between the left camera and the infrared thermal imager; Step 2: Point cloud temperature field reconstruction: The binocular camera obtains the surface speckle image of the measured object and solves the three-dimensional point cloud of the measured object surface; the plane features in the three-dimensional point cloud are used to update the homography matrix H; according to the intrinsic parameter matrix and the updated homography matrix H, the temperature information in the thermal image is mapped to the point cloud corresponding to the visible light coordinates of the left camera, completing the fusion of the temperature field and the three-dimensional point cloud.

2. The method according to claim 1, characterized in that In step 1, according to the corresponding points in the visible light image of the marker points obtained by the left camera and the thermal image, the homography matrix H between the left camera and the infrared thermal imager is solved as follows: the homography matrix H is obtained by fitting the plane using the least squares method.

3. The method according to claim 1, characterized in that In step 1, based on the homography matrix H and the plane features, combined with the intrinsic parameter matrices of the left camera and the infrared thermal imager, the extrinsic parameter matrix between the left camera and the infrared thermal imager is solved as follows: The coordinates of a point in space P = [X, Y, Z] T , the pixel coordinates in the left camera image are [u1, v1], and the pixel coordinates in the thermal image are [u0, v0]; written as homogeneous coordinates p1 = [u1, v1, 1] T, p0 = [u0, v0, 1] T; According to binocular stereo vision, the relationship between pixel coordinates and world coordinates is as follows: p1=K1P,p0=K0(RP+t) Among them, K1 and K0 are the intrinsic parameter matrices of the left camera and the infrared thermal imager respectively; R and t are the rotation matrix and translation vector in the extrinsic parameter matrix between the left camera and the infrared thermal imager respectively; In the world coordinate system, the target plane equation is expressed as: n T P-d=0 The homography transformation from point p1 to p0 is described as p0=Hp1. Combining the expressions of p0 and p1 and the target plane equation, the expression of the homography transformation matrix H is as follows: The 3D point cloud obtained by the binocular camera is used to calculate the calibration plane normal vector n = [n x ,n y ,n z ] T Substitute the distance d from the plane to the origin of the coordinate system into the H expression (1) and solve to obtain the transformation matrix R and t from the infrared thermal imager to the left camera.

4. The method for rapid reconstruction of binocular visual point cloud temperature field based on homography transformation as claimed in claim 3, characterized in that: In step 2, according to the rotation matrix R, the translation vector t and the updated homography matrix H, the temperature information in the thermal image is mapped to the point cloud corresponding to the visible light coordinates, which is: The mapping relationship between the thermal image coordinate p0 and the left camera visible light coordinate p1 is: According to the above formula, the temperature information in the thermal image is first mapped to the visible light coordinates; then based on the corresponding relationship between the left camera coordinates and the three-dimensional point cloud coordinates, the mapping of the temperature information to the three-dimensional point cloud is realized.

5. The method for rapid reconstruction of binocular visual point cloud temperature field based on homography transformation as claimed in claim 1, characterized in that: In step 2, when mapping, the temperature information of the thermal image is interpolated and calculated, and the temperature information in the thermal image is accurately mapped to the point cloud corresponding to the visible light coordinates of the left camera.

6. The method for rapid reconstruction of binocular visual point cloud temperature field based on homography transformation according to claim 1, characterized in that: The interpolation calculation adopts a bicubic difference algorithm.

7. The method for rapid reconstruction of binocular visual point cloud temperature field based on homography transformation according to claim 1, characterized in that: In step 2, before mapping the temperature information of the thermal image, the thermal image is further subjected to distortion correction, resolution adjustment and / or noise filtering.

8. The method for rapid reconstruction of binocular visual point cloud temperature field based on homography transformation as claimed in claim 1, characterized in that: In step 2, the binocular camera acquires a surface speckle image of the measurement object, and solves the three-dimensional point cloud of the measurement object surface by analyzing the displacement change of the speckle through a three-dimensional digital image correlation 3D-DIC method to solve the three-dimensional point cloud of the measurement object surface.

9. The method for rapid reconstruction of binocular visual point cloud temperature field based on homography transformation as claimed in claim 1, characterized in that: The optical axes of the left camera and the infrared thermal imager are parallel.

Images