Heliostat surface shape detection method and detection system

Abstract

The invention discloses a heliostat surface type detection method and detection system, belongs to the technical field of heliostat surface type detection, and can solve the problems of low detection precision and low detection efficiency of an existing detection mode. The method comprises the following steps: S1, collecting a plurality of fringe images with different frequencies reflected by the heliostat; s2, extracting phase data information of each stripe image; and S3, performing three-dimensional reconstruction on the reflecting surface of the heliostat according to the phase data information of the plurality of stripe images to obtain a fitting curved surface of the heliostat, and determining a surface type detection result of the heliostat according to the fitting curved surface. The method is used for detecting the surface type of the heliostat reflecting surface.

Description

Technical Field

[0001] The invention relates to a heliostat mirror surface type detection method and a detection system, belonging to the technical field of heliostat mirror surface type detection. Background Art

[0002] Heliostats are a key component of tower solar power generation technology, and the accuracy of their curvature is directly related to the efficiency of light collection. There are two existing methods for detecting the reflective surface of heliostats: one is to detect heliostats using a laser beam deflection scanning method, but this method has low detection accuracy. The other is to detect heliostats using a secondary reflector surface quality detection method based on image recognition, but this method has low detection efficiency. Summary of the invention

[0003] The present invention provides a heliostat mirror type detection method and detection system, which can solve the problems of low detection accuracy and low detection efficiency of the existing detection method.

[0004] In one aspect, the present invention provides a method for detecting a heliostat surface type, the method comprising:

[0005] S1, collecting multiple fringe images of different frequencies reflected by the heliostat;

[0006] S2, extracting phase data information of each fringe image;

[0007] S3. Reconstruct the reflection surface of the heliostat in three dimensions according to the phase data information of the multiple fringe images to obtain a fitting surface of the heliostat, and determine a surface shape detection result of the heliostat according to the fitting surface.

[0008] Optionally, the extracting of phase data information of each fringe image in S2 is specifically:

[0009] A deep neural network is used to decouple multiple fringe images separately to obtain the phase data information of each fringe image.

[0010] Optionally, the loss function of the deep neural network is a four-element composite loss function.

[0011] Optionally, the four-element composite loss function consists of phase angle loss, difference frequency consistency loss, adversarial reflection suppression loss and structural similarity loss.

[0012] Optionally, the step of performing three-dimensional reconstruction on the reflection surface of the heliostat according to the phase data information of the plurality of fringe images to obtain the fitting surface of the heliostat in S3 specifically includes:

[0013] Generate a depth map based on phase data information of multiple fringe images;

[0014] The depth map is used to perform three-dimensional reconstruction on the reflection surface of the heliostat to obtain a fitting curved surface of the heliostat.

[0015] Optionally, the three-dimensionally reconstructing the reflection surface of the heliostat by using the depth map to obtain the fitting surface of the heliostat specifically includes:

[0016] Converting the depth map into three-dimensional point cloud data using a camera imaging model;

[0017] The three-dimensional point cloud data is reconstructed using a volume fusion algorithm to obtain a fitting surface of the heliostat.

[0018] On the other hand, the present invention provides a heliostat surface type detection system based on any one of the above heliostat surface type detection methods, the system comprising:

[0019] A collection unit, used for collecting multiple fringe images of different frequencies reflected by the heliostat;

[0020] An analysis unit, used for extracting phase data information of each fringe image;

[0021] The reconstruction unit performs three-dimensional reconstruction on the reflection surface of the heliostat according to the phase data information of the multiple fringe images to obtain the fitting surface of the heliostat, and determines the surface shape detection result of the heliostat according to the fitting surface.

[0022] Optionally, the acquisition unit includes:

[0023] A transport module, used for transporting the heliostat to a detection position;

[0024] A canvas, arranged directly opposite to the detection position, having a plurality of stripe patterns with different phases;

[0025] Two scroll modules are respectively connected to two ends of the canvas, and are used to drive the canvas to move so as to sequentially display the stripe patterns on the canvas;

[0026] The acquisition module is arranged on one side of the canvas and is used for acquiring a plurality of fringe images reflected by the heliostat.

[0027] Optionally, the acquisition unit further includes:

[0028] The supporting module is arranged below the canvas and is used to support the canvas so that the stripe pattern on the canvas is in a flattened state.

[0029] Optionally, the acquisition module includes:

[0030] An image collector, disposed on one side of the canvas, for collecting a plurality of fringe images reflected by the heliostat;

[0031] The adjusting structure is connected to the image collector and is used to adjust the height of the image collector so that the image collector is located within the focal length of the heliostat.

[0032] The beneficial effects that the present invention can produce include:

[0033] The heliostat surface type detection method provided by the present invention extracts the phase data information of each fringe image in multiple fringe images of different frequencies, and performs three-dimensional reconstruction of the reflecting surface based on the multiple phase data information. Through the frequency difference, the real morphological information of the reflecting surface can be accurately extracted, thereby effectively eliminating the influence of parasitic reflection and surface morphology, and reducing the error caused by single frequency fringes.

[0034] The heliostat mirror type detection method provided by the present invention uses a four-element composite loss function that couples geometric constraints with data-driven constraints when extracting phase data information of each fringe image using a deep neural network. The loss function can avoid the failure problem of mutation penalty at the boundary of the traditional loss function, suppress the abnormal phase shift caused by parasitic reflection, and effectively suppress the step artifacts in the phase unfolding, thereby improving the smoothness of the surface.

[0035] The heliostat surface type detection system provided by the present invention controls the canvas to display the stripe pattern through the scroll module, collects multiple stripe images reflected by the heliostat using the acquisition module, and then uses the phase shift method to detect the mirror surface type. This method improves the detection accuracy and detection speed of the heliostat surface type detection and greatly reduces the detection cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 A flow chart of a heliostat type detection method provided by an embodiment of the present invention;

[0037] Figure 2 A schematic diagram of the three-dimensional structure of a heliostat mirror type detection system provided by an embodiment of the present invention;

[0038] Figure 3 This is a schematic diagram of the planar structure of a heliostat detection system provided by an embodiment of the present invention.

[0039] Reference numerals:

[0040] 1. First scroll; 2. Second scroll; 3. First drive motor; 4. Second drive motor; 5. Retractable rod; 6. Support platform; 7. Camera; 8. Heliostat; 9. Transmission module; 10. Moving crossbar; 11. Canvas; 12. Smooth structure; 13. Computer; 14. Vertical rod. DETAILED DESCRIPTION

[0041] The present invention is described in detail below in conjunction with embodiments, but the present invention is not limited to these embodiments.

[0042] The embodiment of the present invention provides a heliostat mirror type detection method, such as Figures 1 to 3 As shown, the method includes:

[0043] S1. Collect multiple fringe images of different frequencies reflected by the heliostat 8.

[0044] In practical applications, a collection device can be used to collect multiple fringe images of different frequencies reflected by the heliostat 8. The embodiment of the present invention does not limit the specific structure of the collection device, and the collection device can be an existing fringe image collection structure or a collection unit provided by the detection system of the present invention. The specific structure of the collection unit will be described in detail when describing the detection system of the present invention.

[0045] In the present invention, different fringe images have different frequencies.

[0046] S2. Extract the phase data information of each fringe image.

[0047] Specifically: a deep neural network is used to decouple multiple fringe images separately to obtain the phase data information of each fringe image.

[0048] The loss function of this deep neural network is a four-element composite loss function that couples geometric constraints with data-driven constraints. The specific loss architecture is:

[0049] Assume that the network input contains The fringe image group of frequencies , the output is the phase prediction value corresponding to each frequency , is the true value of the phase, then the loss function is defined as a weighted combination of the following four parts:

[0050] Part I: ;

[0051] The phase angle loss is a period-sensitive angle residual calculation for the phase prediction error of each frequency. This loss forces the network to accurately predict the main phase value and uses cosine distance instead of mean square error to avoid the failure of mutation penalty at the boundary of traditional loss. In each training batch, the phase angle loss of all samples and all frequencies is calculated. The loss value is back-propagated to update the network weights, giving priority to minimizing the local phase error.

[0052] Part II:

[0053] ;

[0054] is the difference frequency consistency loss, in the above formula, To calibrate the actual thickness information of the plate, It is a modular operation to ensure that the phase difference is always within within the cycle; For The number of combinations of 2 frequencies taken from different frequencies, As the denominator, it is used to normalize the phase difference calculation results of all frequencies so that the loss value can be measured at an appropriate scale to avoid the loss value being too large or too small due to the influence of frequency on the quantity, thereby more reasonably reflecting the overall difference between the predicted phase difference of different frequencies and the theoretical value. The constraint conditions are established using the theoretical relationship between the phase differences of multiple frequencies. Assuming that any two frequencies actually satisfy ( is the ratio coefficient of the ith frequency to the jth frequency), the predicted phase difference should be consistent with the theoretical difference. This loss function constrains the predicted phase difference values ​​of different frequencies to be consistent with their theoretical values, avoiding local phase mutations caused by parasitic reflections, thereby improving global phase consistency, and calculating the loss for all possible frequency combinations after each forward propagation to ensure that the multi-frequency phase output by the network is consistent with the actual phase.

[0055] Part III: ;

[0056] is the antagonistic reflection suppression loss, in the above formula, The generator (main network) receives the input stripe image I and outputs the predicted phase. D is the discriminator network, which takes the phase map as input and outputs [ 0,1 ] The probability value of indicates whether the input comes from the real data distribution without reflection interference. In the calculation, it can be approximated by the average value of the batch samples. Construct an auxiliary discriminator to distinguish the parasitic reflection artifacts in the predicted phase image, and the generator The loss forces the network to generate results consistent with the true non-reflection interference phase distribution, especially improving the phase continuity in the highly reflective area.

[0057] Part 4: ;

[0058] is the structural similarity loss. In the above formula, is the phase gradient (computed via the Sobel operator while preserving the gradients in both x and y directions). is the mean of the phase gradients in the current batch, is the variance of the phase gradient, is the covariance between the predicted phase gradient and the true gradient, is a stability constant. This loss function introduces structural similarity constraints on phase gradients to maintain the local structural features of the predicted phase image. This loss effectively suppresses step artifacts in phase unwrapping and improves surface smoothness.

[0059] The final composite loss function is a linear combination of the above four items:

[0060] ;

[0061] In the above formula, the coefficient Obtained through model training. For example, after training, the coefficient values ​​are: .

[0062] The four-element composite loss function It avoids the failure problem of mutation penalty at the boundary of traditional loss function, suppresses abnormal phase shift caused by parasitic reflection, effectively suppresses step artifacts in phase unfolding, and improves surface smoothness.

[0063] After acquiring multiple stripe images, a deep neural network is used for stripe decoupling to recover the original phase data information from the stripe images containing multiple frequencies and angles. In order to improve the generalization ability of the network, each stripe image will undergo data enhancement processing before decoupling, such as rotation, cropping and illumination changes. Stripe images usually cause phase misalignment due to factors such as reflection and distortion. The goal of the network is to extract clear phase data information from them. The network structure contains multiple convolutional layers, focusing on extracting spatial features in stripe images, and retaining more detail information by introducing jump connections to enhance the processing ability of stripe images. Finally, the decoupled phase image is restored by deconvolution operation. In terms of loss function design, the present invention measures the error between the output phase and the true phase by designing a four-element composite loss function. The trained deep neural network can effectively decouple phase data information from complex stripe images and provide accurate phase data for subsequent three-dimensional reconstruction.

[0064] S3. Reconstruct the reflection surface of the heliostat 8 in three dimensions according to the phase data information of the multiple fringe images to obtain a fitting surface of the heliostat 8 , and determine the surface shape detection result of the heliostat 8 according to the fitting surface.

[0065] The three-dimensional reconstruction of the reflection surface of the heliostat 8 is performed according to the phase data information of the multiple fringe images to obtain the fitting surface of the heliostat 8, which specifically includes:

[0066] (1) Generate a depth map based on the phase data information of multiple fringe images.

[0067] The mapping model that converts phase data information into depth values ​​in space through triangulation is as follows:

[0068] ;

[0069] in, is the phase data information; is the baseline distance from the camera to the canvas; is the phase period wavelength; is the camera focal length; is the calibration correction item.

[0070] Substitute the absolute phase value of each pixel in the phase data information into the mapping model, calculate the corresponding depth value, and generate an initial depth map. Then, fill in the missing area through a post-processing interpolation algorithm (such as the nearest neighbor interpolation method) and output an optimized depth map.

[0071] (2) The reflection surface of the heliostat 8 is three-dimensionally reconstructed using the depth map to obtain a fitting surface of the heliostat 8.

[0072] Specifically: firstly, the depth map is converted into three-dimensional point cloud data using a camera imaging model; then, the surface of the three-dimensional point cloud data is reconstructed using a volume fusion algorithm to obtain a fitting surface of the heliostat 8 .

[0073] Since stripe images of different frequencies can provide different depth information, they can help accurately reconstruct the three-dimensional surface. First, the camera imaging model is used to convert the depth value of each pixel in the depth map into a three-dimensional space coordinate to obtain three-dimensional point cloud data; then, a machine learning model such as a convolutional neural network (CNN) or a generative adversarial network (GAN) is used to repair the noise of the three-dimensional point cloud data to obtain optimized three-dimensional point cloud data. Finally, the optimized three-dimensional point cloud data is converted into a complete three-dimensional surface by using a volume fusion algorithm (Poisson surface reconstruction). The final three-dimensional reconstruction result is smoothed and hole repaired to generate a fine fitting surface.

[0074] Another embodiment of the present invention provides a heliostat surface type detection system based on any one of the above heliostat surface type detection methods, the system comprising:

[0075] The acquisition unit is used to acquire multiple fringe images of different frequencies reflected by the heliostat 8 .

[0076] The analysis unit is used to extract phase data information of each fringe image.

[0077] The reconstruction unit reconstructs the reflection surface of the heliostat 8 in three dimensions according to the phase data information of the multiple fringe images to obtain a fitting surface of the heliostat 8 and determines a surface shape detection result of the heliostat 8 according to the fitting surface.

[0078] refer to Figure 2 and Figure 3 As shown, the acquisition unit includes:

[0079] The transport module 9 is used to transport the heliostat 8 to the detection position.

[0080] The conveying module 9 conveys the produced heliostat 8 to the detection position. The embodiment of the present invention does not limit the specific structure of the conveying module 9, as long as it can convey the heliostat 8 to the detection position. In practical applications, the conveying module 9 can use a structure such as a conveyor belt.

[0081] The canvas 11 is arranged opposite to the detection position and has a plurality of stripe patterns with different phases.

[0082] The canvas 11 contains a series of stripe patterns with different phases, and the interval between two adjacent stripe patterns is generally 40 cm to 60 cm. In practical applications, the interval between adjacent stripe patterns can be set to 50 cm.

[0083] The two scroll modules are respectively connected to the two ends of the canvas 11 and are used to drive the canvas 11 to move so as to display the stripe patterns on the canvas 11 in sequence.

[0084] Specifically, the scroll module includes:

[0085] A reel, on which the end of the canvas 11 is rolled;

[0086] The driving motor is used to drive the reel to rotate so as to roll the canvas 11 onto the reel.

[0087] In an embodiment of the present invention, the scroll in the scroll module connected to the right end of the canvas 11 is recorded as the first scroll 1, and the corresponding driving motor is recorded as the first driving motor 3; the scroll in the scroll module connected to the left end of the canvas 11 is recorded as the second scroll 2, and the corresponding driving motor is recorded as the second driving motor 4.

[0088] refer to Figure 2 As shown, the first reel 1 stretches a group of striped canvases 11 in the second reel 2 and displays them on the support table 6. The second reel 2 shrinks the striped canvases 11 in the first reel 1 and retracts them into the second reel 2. The first drive motor 3 provides power for the first reel 1. The second drive motor 4 provides power for the second reel 2.

[0089] The acquisition module is arranged on one side of the canvas 11 and is used to acquire a plurality of fringe images reflected by the heliostat 8 .

[0090] Specifically, the acquisition module includes:

[0091] The image collector is arranged on one side of the canvas 11 and is used to collect multiple stripe images reflected by the heliostat 8 ; in practical applications, the image collector may be a camera 7 .

[0092] The heliostat 8 is the object to be detected. The camera 7 is used to collect the fringe image reflected by the heliostat 8.

[0093] The adjustment structure is connected to the image collector and is used to adjust the height of the image collector so that the image collector is located within the focal length of the heliostat 8 .

[0094] The regulatory structure includes:

[0095] The bracket includes two vertical rods 14, and the two vertical rods 14 have guide grooves along their length directions on the opposite surfaces thereof;

[0096] The two ends of the movable cross bar 10 are respectively arranged in two guide grooves and can move along the guide grooves; the image collector is connected to the movable cross bar 10 .

[0097] The movable cross bar 10 can realize vertical movement, so that the position of the camera 7 can be adjusted.

[0098] Furthermore, the acquisition unit also includes:

[0099] The supporting module is arranged below the canvas 11 and is used to support the canvas 11 so that the stripe pattern on the canvas 11 is in a flat state.

[0100] In an embodiment of the present invention, the support module includes:

[0101] A support platform 6 is provided below the canvas 11 and is used to support the canvas 11;

[0102] A plurality of retractable rods 5 are disposed below the support platform 6 and are used to adjust the height of the support platform 6 so that the striped pattern on the canvas 11 is in a flattened state.

[0103] The support table 6 supports the striped canvas 11 and provides a placement platform for the striped canvas 11 .

[0104] The retractable rod 5 provides support for the support platform 6 and can be retracted to adjust the height of the support platform 6 .

[0105] Preferably, the acquisition unit further includes:

[0106] The smooth structure 12 is installed at the edge where the support platform 6 contacts the canvas 11 .

[0107] The smooth structure 12 can enable the canvas 11 to run smoothly between the two reels and reduce the wear on the canvas 11. The embodiment of the present invention does not limit the specific structure of the smooth structure 12. In practical applications, the smooth structure 12 can be a structure such as a rounded corner strip.

[0108] The working process of the heliostat mirror detection system is described in detail below.

[0109] First, the acquisition unit is placed behind the heliostat production line. After the heliostat 8 is manufactured, the heliostat 8 is brought to the detection position by the transmission module 9. The position of the heliostat 8 is fixed and the camera 7 is adjusted up and down by moving the crossbar 10 so that it is located within the focal length of the heliostat 8, and the camera 7 is calibrated to ensure the accuracy of the detection. The present invention uses an intelligent calibration method based on deep learning, which is a new method for calibrating the camera 7 using deep learning technology. At the beginning of this intelligent calibration method, a large amount of image data needs to be collected and preprocessed. After preprocessing, the collected data is trained using a convolutional neural network. Through continuous iteration and optimization, the model can more accurately estimate the internal parameters (such as focal length, principal point position) and external parameters (such as rotation and translation matrix) of the camera 7, as well as the distortion coefficient. The trained model can be applied to the calibration of the camera 7 in the heliostat reflective surface detection device. By acquiring the image collected by the camera 7, the model can quickly and accurately estimate the internal and external parameters of the camera 7 to achieve the calibration of the camera 7.

[0110] After the camera 7 is calibrated, the heliostat surface type detection part begins. The first drive motor 3 drives the first reel 1 to roll the canvas 11 inside the second reel 2, wherein the canvas 11 has a group of 16 stripe patterns printed in sequence. The spacing between each stripe pattern is 50 cm, and the spacing of 50 cm is set to prevent the influence between adjacent stripe patterns. In order to reduce the deformation of the canvas 11 as much as possible, the support platform 6 is set up with a telescopic rod 5 so that the canvas 11 is in a state of tightly fitting the support platform 6. At the same time, under the action of the smooth structure 12, the canvas 11 can run smoothly between the two reels and reduce the wear on the canvas 11. Every time the canvas 11 displays a stripe pattern, the camera 7 is used to capture the image of the stripe pattern reflected by the heliostat 8 and save the image to the computer 13. Repeat 16 times to complete the collection of a group of stripe patterns. After a group of stripe patterns are collected, the second drive motor 4 drives the second reel 2 to roll the canvas 11 inside the first reel 1 so that the collection can continue next time.

[0111] After obtaining a set of fringe images containing the surface shape information of the reflective surface of the heliostat 8, the internal code of the computer 13 is used to analyze the set of fringe images. Then, the surface shape of the reflective surface of the heliostat 8 is obtained by code fitting. If the fitted reflective surface shape is within the qualified deviation range from the set standard surface shape, the heliostat 8 is considered qualified and shipped out of the factory. If the fitted reflective surface shape is significantly different from the set standard surface shape, it is returned for correction. After correction, it is put into the surface shape detection system again for detection, and this is repeated for multiple times until the heliostat 8 is qualified.

[0112] The functions implemented by the internal code in the computer 13 are the functions implemented by the analysis unit and the reconstruction unit, which mainly solve the phase recovery and three-dimensional reconstruction problems in the fringe image. Specifically, the fringe image is first decoupled using a deep neural network to extract accurate phase data information from it. Next, combined with machine learning optimization technology, the decoupled phase data information is converted into a depth map, and the depth information in the depth map is optimized and reconstructed to generate a high-precision three-dimensional surface model, thereby obtaining the fitting surface of the heliostat 8. In the whole process, the combination of neural networks and multi-frequency methods not only ensures high-precision phase recovery, but also effectively eliminates complex reflection interference, and finally achieves high-precision three-dimensional reconstruction based on machine learning.

[0113] After the detection starts, the present invention does not require multiple steps and can complete the detection of the reflective surface of the heliostat 8 in a short time.

[0114] The present invention uses a scroll control canvas 11 to replace the display screen to play the stripe pattern, which greatly reduces the cost.

[0115] The present invention uses a phase shift method to detect the mirror surface, thereby improving the detection accuracy and detection speed of the heliostat 8.

[0116] The present invention uses an independently innovatively designed detection architecture, which can quickly interact with the heliostat production line, thereby shortening the overall operation time of the entire production line and the detection device.

[0117] The present invention decouples the complex phase data information in the fringe image through a deep neural network, ensuring that accurate phase data can be recovered from noise and parasitic reflections.

[0118] The present invention converts phase information into an accurate depth map, further improves the accuracy of 3D reconstruction through multi-frequency optimization, and uses a machine learning algorithm to enhance the fineness and stability of 3D reconstruction.

[0119] The above are only a few embodiments of the present application and do not constitute any form of limitation to the present application. Although the present application is disclosed as above with preferred embodiments, it is not intended to limit the present application. Any technician familiar with the profession, without departing from the scope of the technical solution of the present application, using the technical content disclosed above to make slight changes or modifications are equivalent to equivalent implementation cases and fall within the scope of the technical solution.

Claims

1. A heliostat mirror type detection method, characterized in that: The method comprises: S1, collecting multiple fringe images of different frequencies reflected by the heliostat; S2, extracting phase data information of each fringe image; S3. Reconstruct the reflection surface of the heliostat in three dimensions according to the phase data information of the multiple fringe images to obtain a fitting surface of the heliostat, and determine a surface shape detection result of the heliostat according to the fitting surface.

2. The method according to claim 1, characterized in that: The extraction of phase data information of each fringe image in S2 is specifically: A deep neural network is used to decouple multiple fringe images separately to obtain the phase data information of each fringe image.

3. The method according to claim 2, characterized in that The loss function of the deep neural network is a four-element composite loss function.

4. The method according to claim 3, characterized in that: The four-element composite loss function is composed of phase angle loss, difference frequency consistency loss, adversarial reflection suppression loss and structural similarity loss.

5. The method according to claim 1, characterized in that The step S3 of performing three-dimensional reconstruction on the reflection surface of the heliostat according to the phase data information of the plurality of fringe images to obtain the fitting surface of the heliostat specifically includes: Generate a depth map based on phase data information of multiple fringe images; The depth map is used to perform three-dimensional reconstruction on the reflection surface of the heliostat to obtain a fitting curved surface of the heliostat.

6. The method according to claim 5, characterized in that The three-dimensional reconstruction of the reflection surface of the heliostat by using the depth map to obtain the fitting surface of the heliostat specifically includes: Converting the depth map into three-dimensional point cloud data using a camera imaging model; The three-dimensional point cloud data is reconstructed using a volume fusion algorithm to obtain a fitting surface of the heliostat.

7. A heliostat surface type detection system based on the heliostat surface type detection method according to any one of claims 1 to 6, characterized in that: The system comprises: A collection unit, used for collecting multiple fringe images of different frequencies reflected by the heliostat; An analysis unit, used for extracting phase data information of each fringe image; The reconstruction unit performs three-dimensional reconstruction on the reflection surface of the heliostat according to the phase data information of the multiple fringe images to obtain the fitting surface of the heliostat, and determines the surface shape detection result of the heliostat according to the fitting surface.

8. The system according to claim 7, characterized in that The acquisition unit comprises: A transport module, used for transporting the heliostat to a detection position; A canvas, arranged directly opposite to the detection position, having a plurality of stripe patterns with different phases; Two scroll modules are respectively connected to two ends of the canvas, and are used to drive the canvas to move so as to sequentially display the stripe patterns on the canvas; The acquisition module is arranged on one side of the canvas and is used for acquiring a plurality of fringe images reflected by the heliostat.

9. The system according to claim 8, characterized in that The acquisition unit also includes: The supporting module is arranged below the canvas and is used to support the canvas so that the stripe pattern on the canvas is in a flattened state.

10. The system according to claim 8, characterized in that The acquisition module comprises: An image collector, disposed on one side of the canvas, for collecting a plurality of fringe images reflected by the heliostat; The adjusting structure is connected to the image collector and is used to adjust the height of the image collector so that the image collector is located within the focal length of the heliostat.

Images