WO2023023961A1 - 一种基于激光线阵的piv图像标定装置及方法 - Google Patents
一种基于激光线阵的piv图像标定装置及方法 Download PDFInfo
- Publication number
- WO2023023961A1 WO2023023961A1 PCT/CN2021/114416 CN2021114416W WO2023023961A1 WO 2023023961 A1 WO2023023961 A1 WO 2023023961A1 CN 2021114416 W CN2021114416 W CN 2021114416W WO 2023023961 A1 WO2023023961 A1 WO 2023023961A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- laser
- image
- laser line
- distortion
- grating
- Prior art date
Links
- 238000000034 method Methods 0.000 title claims abstract description 28
- 238000013528 artificial neural network Methods 0.000 claims abstract description 29
- 230000003287 optical effect Effects 0.000 claims abstract description 29
- 238000002474 experimental method Methods 0.000 claims abstract description 23
- 238000011084 recovery Methods 0.000 claims abstract description 10
- 238000012545 processing Methods 0.000 claims abstract description 4
- 238000003062 neural network model Methods 0.000 claims description 30
- 238000009434 installation Methods 0.000 claims description 29
- 238000001514 detection method Methods 0.000 claims description 11
- 230000008439 repair process Effects 0.000 claims description 8
- 210000002569 neuron Anatomy 0.000 claims description 4
- 230000004913 activation Effects 0.000 claims description 3
- 238000000605 extraction Methods 0.000 claims description 2
- 230000035939 shock Effects 0.000 abstract description 18
- 238000012937 correction Methods 0.000 abstract description 2
- 238000000827 velocimetry Methods 0.000 abstract 1
- 238000000917 particle-image velocimetry Methods 0.000 description 29
- 238000012549 training Methods 0.000 description 10
- 230000000694 effects Effects 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 230000006870 function Effects 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 4
- 238000005259 measurement Methods 0.000 description 4
- 239000002245 particle Substances 0.000 description 4
- 239000011159 matrix material Substances 0.000 description 3
- 238000005457 optimization Methods 0.000 description 3
- 239000000243 solution Substances 0.000 description 3
- PXHVJJICTQNCMI-UHFFFAOYSA-N Nickel Chemical compound [Ni] PXHVJJICTQNCMI-UHFFFAOYSA-N 0.000 description 2
- 238000003491 array Methods 0.000 description 2
- 238000005311 autocorrelation function Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 239000000700 radioactive tracer Substances 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 238000012795 verification Methods 0.000 description 2
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 239000012482 calibration solution Substances 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 229910052759 nickel Inorganic materials 0.000 description 1
- 238000007747 plating Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000001454 recorded image Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M9/00—Aerodynamic testing; Arrangements in or on wind tunnels
- G01M9/02—Wind tunnels
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M9/00—Aerodynamic testing; Arrangements in or on wind tunnels
- G01M9/08—Aerodynamic models
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01P—MEASURING LINEAR OR ANGULAR SPEED, ACCELERATION, DECELERATION, OR SHOCK; INDICATING PRESENCE, ABSENCE, OR DIRECTION, OF MOVEMENT
- G01P21/00—Testing or calibrating of apparatus or devices covered by the preceding groups
- G01P21/02—Testing or calibrating of apparatus or devices covered by the preceding groups of speedometers
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01P—MEASURING LINEAR OR ANGULAR SPEED, ACCELERATION, DECELERATION, OR SHOCK; INDICATING PRESENCE, ABSENCE, OR DIRECTION, OF MOVEMENT
- G01P5/00—Measuring speed of fluids, e.g. of air stream; Measuring speed of bodies relative to fluids, e.g. of ship, of aircraft
- G01P5/18—Measuring speed of fluids, e.g. of air stream; Measuring speed of bodies relative to fluids, e.g. of ship, of aircraft by measuring the time taken to traverse a fixed distance
- G01P5/20—Measuring speed of fluids, e.g. of air stream; Measuring speed of bodies relative to fluids, e.g. of ship, of aircraft by measuring the time taken to traverse a fixed distance using particles entrained by a fluid stream
Definitions
- the invention relates to the technical field of laser velocity measurement and image restoration, in particular to a PIV image calibration device and method based on a laser line array.
- Particle Image Velocimetry (PIV, Particle Image Velocimetry) technology is a non-contact flow velocity field optical measurement technology. The measurement of the motion velocity in the flow velocity field characterized by microclusters, and is widely used in wind tunnel experiments.
- the image spatial resolution needs to be calibrated.
- the black and white checkerboard target is placed in the measured area, and after being collected by the camera, the mathematical model of the distortion is solved by using the recorded corner coordinates, and the spatial resolution of the image is obtained everywhere.
- shock waves will be generated locally, and the shock waves will cause local optical diffraction in the measured area.
- the recorded particle trajectory image is distorted, and a more complex distortion effect is produced here, so that the spatial resolution of the image everywhere cannot be accurately obtained.
- the target If the target is placed in the measured area, if it is in a static and windless condition, the shock wave cannot be obtained, and distortion cannot be generated, so that the spatial resolution of the image everywhere cannot be obtained; if it is in a windy condition, the target itself Shock waves will be caused again, and more complex distortion effects will be produced, and the spatial resolution of the image everywhere cannot be accurately obtained.
- none of the above methods can obtain the shock wave distortion image in the real model experiment, and thus cannot obtain the spatial resolution of the image everywhere.
- the purpose of the present invention is to provide a PIV image calibration device and method based on a laser line array, so as to achieve the purpose of accurately obtaining shock wave distortion images during real model experiments.
- the present invention provides the following scheme:
- a PIV image calibration device based on a laser line array comprising:
- the laser emitting part is used to emit a laser line array with equidistant characteristics to form a laser line array optical path;
- An optical component which is used to perform spectroscopic processing on the laser lines in the optical path of the laser line array to form a laser grating in the experimental observation area;
- the camera is used to obtain the distorted laser grating image when the working condition of the wind tunnel experimental section model is adjusted to the PIV experimental working condition; the experimental observation area is located above the wind tunnel experimental section model;
- the background processor is used for the calibration algorithm of distortion recovery based on the neural network, and performs calibration and repair on the distorted laser grating image to obtain the reconstructed laser grating image.
- the laser emitting part includes a fixing frame and a plurality of laser pointers
- a plurality of laser pointers are installed on the fixing frame in a parallel arrangement, and the distance between any two adjacent laser pointers is equal;
- all the laser lines in the emitted laser line array are parallel and coplanar by adjusting the installation angle of the laser emitting part and the distance between two adjacent laser pointers.
- the optical component includes a half mirror and a total reflection mirror sequentially arranged on the optical path of the laser line array;
- the first reflected laser lines and the second reflected laser lines intersect within the experimental observation region to form a staggered laser grating.
- the camera is used to obtain the distorted laser grating image when the working condition of the model in the wind tunnel experiment section is adjusted to the PIV experimental working condition;
- the camera Before turning on the climax wind tunnel, the camera is used to acquire laser grating images before distortion.
- the background processor specifically includes:
- the distorted two-dimensional corner point coordinate information extraction module is used to adopt the corner point detection algorithm to extract the two-dimensional corner point coordinate information of the laser grating on the laser grating image after the distortion;
- the two-dimensional corner point coordinate information determination module before distortion is used for determining the two-dimensional corner point coordinate information before distortion based on the neural network model and the two-dimensional corner point coordinate information of the laser grating on the laser grating image after distortion;
- the reconstruction module is used to obtain the reconstructed laser grating image based on the two-dimensional corner point coordinate information before the distortion.
- the neural network model is a three-layer neural network model; the loss function of the neural network model is a mean square error function; the neural network model includes two hidden layers, and each neuron is added ReLU activation layer;
- the input of the neural network model is the two-dimensional corner coordinate information of the laser grating on the laser grating image after distortion
- the output of the neural network model is the corresponding two-dimensional corner coordinate information before distortion
- a PIV image calibration method based on laser line array comprising:
- the laser grating is a grating formed in the experimental observation area after the laser line in the laser line array light path is spectroscopically processed;
- a neural network-based distortion recovery calibration algorithm is used to perform calibration repair on the distorted laser grating image to obtain a reconstructed laser grating image.
- the laser emitting part includes a fixed frame and a plurality of laser pointers, and the plurality of laser pointers are installed on the fixed frame in a parallel arrangement, and the distance between any two adjacent laser pointers is equal;
- all the laser lines in the emitted laser line array are parallel and coplanar by adjusting the installation angle of the laser emitting part and the distance between two adjacent laser pointers.
- the first reflected laser lines and the second reflected laser lines intersect within the experimental observation region to form a staggered laser grating.
- the neural network-based distortion recovery calibration algorithm performs calibration and repair on the distorted laser grating image to obtain a reconstructed laser grating image, which specifically includes:
- the reconstructed laser grating image is obtained.
- the invention discloses the following technical effects:
- the invention provides a PIV image calibration device and method based on a laser line array.
- the invention collects laser gratings instead of traditional physical targets, and realizes real model experimental calibration in the presence of shock waves without disturbing the real flow field.
- the distortion recovery calibration algorithm based on neural network is used to calibrate and recover the distortion of the laser grating image, so as to achieve the purpose of accurately obtaining the shock wave distortion image in the real model experiment, and then obtain the image space of each place resolution.
- Fig. 1 is the structural representation of a kind of PIV image calibration device based on laser linear array of the present invention
- Fig. 2 is a schematic flow chart of a PIV image calibration method based on a laser line array in the present invention
- Fig. 3 is a kind of flow chart of the image distortion calibration method based on laser line array and neural network of the present invention
- Fig. 4 is the schematic diagram of neural network of the present invention.
- Fig. 5 is the laser grating diagram before and after the distortion of the implementation record of the present invention
- Fig. 5 (a) is the laser grating diagram before the distortion
- Fig. 5 (b) is the laser grating diagram after the distortion
- Fig. 5 (c) is the laser grating after the distortion repair picture.
- An object of the present invention is to provide a new PIV image calibration device without a physical target to replace the traditional physical target, so as to perform real model test calibration in the presence of shock waves without disturbing the real flow field. And on this basis, the neural network algorithm is used to calibrate and restore the distortion of the laser grating image after distortion.
- Another object of the present invention is to provide a software and hardware system for two-dimensional PIV image calibration, to achieve a technical solution and system implementation for obtaining high-precision image calibration without contact and interference.
- This embodiment provides a PIV image calibration device based on a laser line array, which mainly uses a laser line array and optical components to form a laser grating in the experimental observation area as an object-free target.
- the device is used in the PIV experiment of the high Mach number wind tunnel model.
- the schematic diagram of the device is shown in Figure 1:
- the laser emitting part 3 is used to emit a laser line array with equidistant characteristics, so as to form a laser line array optical path.
- the optical component is used for splitting the laser line in the optical path of the laser line array to form a laser grating in the experimental observation area.
- the camera 6 is used to obtain the distorted laser grating image when the working condition of the model 1 in the wind tunnel experiment section is adjusted to the PIV experiment working condition; the experimental observation area is located above the model 1 in the wind tunnel experiment section.
- the background processor is used for calibrating and repairing the distorted laser grating image based on the neural network algorithm to obtain the reconstructed laser grating image.
- the laser line array 3 described in this embodiment includes a fixed frame and a plurality of high-power laser pointers installed on the fixed frame.
- Multiple laser pointers are installed on the fixed frame in a parallel arrangement, and the distance between any two adjacent laser pointers is equal.
- this embodiment uses 10 groups of continuous He-Ne laser pointers with a power of 1W to generate green light with a wavelength of 633nm.
- the optical assembly described in this embodiment includes a half mirror 4 and a total reflection mirror 5 arranged on the optical path of the laser line array. Both the half mirror 4 and the total reflection mirror 5 are located above the model 1 of the wind tunnel experiment section.
- the installation angle of the total reflection mirror 5 is adjusted so that the transmitted laser line is completely reflected after passing through the total reflection mirror 5 to form a second reflected laser line, and then the second reflected laser line is irradiated in the experimental observation area.
- the first reflected laser lines and the second reflected laser lines intersect in the experimental observation area to form a staggered laser grating.
- the materials of the half mirror 4 and the total reflection mirror 5 described in this embodiment are materials such as nickel plating.
- the camera 6 for PIV shooting needs to be arranged; the installation position and installation angle of the camera 6 should be adjusted to ensure that the experimental observation area can be captured.
- the laser grating image captured again by the camera 6 is a distorted laser grating image. Therefore, the camera 6 at this time is used to obtain the distorted laser grating image when the working condition of the model in the wind tunnel experiment section is adjusted to the PIV test working condition.
- the camera 6 Before opening the climax wind tunnel, the camera 6 is used to acquire laser raster images before distortion for subsequent neural network training.
- the camera 6 described in this embodiment is a double-exposure CCD camera.
- the camera 6 is controlled by software to record and shoot the laser grating before and after the wind tunnel operation.
- the camera data is transferred to the background processor for storage through the high-speed Cameralink data line.
- the wind tunnel described in this embodiment is used to generate uniform and stable high-speed airflow, and the experimental model is pre-arranged in the experimental section, and the tracer particles 2 are scattered in the experimental section to enhance the display of the laser grating on the image Effect.
- the distortion recovery calibration algorithm based on the neural network proposed in this embodiment can be used to calibrate the laser grating image after the distortion caused by the shock wave fix.
- corner detection is performed on the two recorded laser grating images, and the corresponding two-dimensional corner coordinate information is extracted. Since the entire spatial position does not change, the corner coordinates detected by the front and back two laser raster images correspond one by one.
- a three-layer neural network structure is constructed, the input is the two-dimensional corner coordinate information extracted from the distorted laser grating image, and the output is the two-dimensional corner coordinate information extracted from the undistorted laser grating image.
- Such a neural network model implies the distortion information of the entire image.
- the laser grating used in this embodiment has no influence on the real flow field, and can record the real distorted image caused by the shock wave.
- This embodiment provides a laser line array-based PIV image calibration method, which is applied to a laser line array-based PIV image calibration device described in Embodiment 1.
- the method includes:
- Step 201 Obtain the distorted laser grating image when the working condition of the wind tunnel experimental section model is adjusted to the PIV experimental working condition; the laser grating is formed in the experimental observation area after performing spectroscopic processing on the laser lines in the laser line array light path raster.
- Step 202 Calibrate and repair the distorted laser grating image based on the neural network distortion recovery calibration algorithm to obtain the reconstructed laser grating image.
- a laser line array-based PIV image calibration method described in this embodiment further includes:
- the laser emitting part includes a fixed frame and a plurality of laser pointers, and the plurality of laser pointers are installed on the fixed frame in a parallel arrangement, and the distance between any two adjacent laser pointers is equal; During operation, all the laser lines in the emitted laser line array are parallel and coplanar by adjusting the installation angle of the laser emitting part and the distance between two adjacent laser pointers.
- a laser line array-based PIV image calibration method described in this embodiment further includes:
- a half mirror and a total reflection mirror are sequentially arranged on the optical path of the laser line array.
- the installation angle of the half-mirror When working, by adjusting the installation angle of the half-mirror, after the laser line passes through the half-mirror, a part of the laser line is transmitted, and the other part of the laser line is reflected to form the first reflected laser line , and then the first reflected laser line is irradiated in the experimental observation area; the installation angle of the total reflection mirror is adjusted so that the transmitted laser line is completely reflected after passing through the total reflection mirror to form a second reflected laser line, Then the second reflected laser line is irradiated in the experimental observation area; the first reflected laser line and the second reflected laser line intersect in the experimental observation area to form a staggered laser grating.
- step 202 specifically includes:
- the corner detection algorithm is used to extract the two-dimensional corner coordinate information of the laser grating on the distorted laser grating image.
- the two-dimensional corner point coordinate information before the distortion is determined.
- the reconstructed laser grating image is obtained.
- This embodiment provides an image distortion calibration method based on laser line array and neural network, please refer to Figure 3, including:
- Step 1 Build the optical path of the laser line array; specifically:
- Step 2 Adjust the optical lens to form a grating; specifically:
- Step 2.1 First arrange a half mirror on the optical path of the laser line array; the half mirror is located above the model of the wind tunnel experiment section, and the half mirror is used to directly transmit a part of the laser line of the laser line array, and transmit the laser A part of the laser line of the line array is reflected; then adjust the installation position and installation angle of the half mirror to make the reflected light illuminate the experimental observation area, that is, the flow observation area.
- Step 2.2 First arrange the total reflection mirror on the optical path of the transmitted light; the total reflection mirror is located above the model of the wind tunnel experiment section, and the total reflection mirror is used to reflect all the transmitted light; secondly, adjust the installation position and installation of the total reflection mirror The angle is such that the reflected and transmitted light also illuminates the experimental observation area, and then intersects with the reflected light passing through the half mirror to form a laser grating.
- Step 3 Collect images before and after distortion; specifically:
- Step 3.1 First set up the camera on the experimental bench; then adjust the installation position and angle of the camera so that the camera can accurately capture the experimental observation area.
- Step 3.2 Spread tracer particles in the wind tunnel experiment section to enhance the reflection effect of the laser grating.
- Step 3.3 Before the wind tunnel is opened, use the camera to record the laser grating image, and the laser grating on the laser grating image is not distorted at this time.
- Step 3.4 Turn on the climax wind tunnel, and adjust the working condition of the wind tunnel experiment section to the PIV experiment working condition. At this time, shock waves are generated on the surface of the model in the wind tunnel experiment section, and the laser grating is distorted. Finally, the camera is used to record the distortion after distortion laser raster image;
- Step 4 Corner detection; specifically:
- the two-dimensional corner coordinate information of the laser grating on the laser grating image before distortion and the laser grating image after distortion is respectively extracted by corner detection algorithm.
- the corner detection algorithm is as follows:
- a window with a set size is used to move in all directions of the laser raster image, and the autocorrelation function of the gray level change in the window is calculated during the movement, as shown in formula (1):
- the corner position When it is detected that R is greater than 0, the corner position can be located and the two-dimensional corner coordinate information can be extracted.
- Step 5 Build the neural network.
- the neural network contains two hidden layers, and a ReLU activation layer is added after each neuron.
- the computational mathematical expression of a neuron is as follows:
- the loss of the neural network is calculated using the mean square error function, that is, the loss function of the neural network is:
- (x, y) is the predicted coordinates output by the neural network
- (x', y') is the real coordinates output by the neural network, that is, the two-dimensional corner coordinates before distortion determined by the corner detection algorithm.
- n is the number of batches in a batch during training.
- Step 6 training to obtain a neural network model; the input of the neural network model is the two-dimensional corner coordinate information after distortion, and the output is the two-dimensional corner coordinate information before distortion.
- the previously detected two-dimensional corner coordinate information before distortion and the two-dimensional corner coordinate information after distortion constitute a training set sample, and a group of 8 samples is used as a batch for neural network weight training.
- the Adam optimization learning algorithm is selected for neural network optimization training.
- the optimization process is as follows:
- Step 6.2 Under the current weight, bring in a batch of samples to calculate the neural network output loss Loss t ( ⁇ t-1 ), and find its gradient to the weight ⁇ .
- the calculation formula is as follows:
- t represents the current iteration number
- t-1 is the previous iteration number
- Step 6.:3 Calculate the biased first-order moment of momentum and the biased second-order moment of momentum, the calculation formula is as follows:
- Step 6.4 Calculate the unbiased first-order moment of momentum and the unbiased second-order moment of momentum, the calculation formula is as follows:
- Step 6.5 Calculate and update the neural network weights, the calculation formula is as follows:
- Step 6.6 Repeat steps 6.2 to 6.5 until the loss does not decrease and the training is considered complete.
- Step 6.7 Use the unused two-dimensional corner point coordinate information as a verification set to verify the accuracy of the prediction of the neural network model. When the accuracy is higher than 90%, the neural network model is considered to meet the requirements.
- Step 7 The calibration is completed, and the image distortion is repaired.
- each pixel point (x, y, I) on the distorted laser grating image where, I is the grayscale information on the pixel.
- the coordinates (x, y) are input into the neural network model containing distortion information, the coordinates (x', y') on the laser raster image before distortion are predicted, and all pixels on the laser raster image after distortion (x, y') are traversed.
- y, I the corresponding pre-distorted laser grating image can be reconstructed, and its effect is shown in Figure 5.
- the invention discloses a PIV image calibration device and method based on a laser line array, including: a laser emitting part emitting a series of equidistant laser line arrays, and a plurality of sets of optical lenses are arranged on the optical path of the laser line array to perform a calibration of the laser line Reflection and projection, that is, to realize the light splitting of the laser line array, and to form a laser grid with the laser line array vertically intersecting near the model of the wind tunnel experiment section by rationally arranging the installation position and installation angle of the optical lens.
- a camera is arranged outside the observation window of the model in the wind tunnel experiment section, and the camera is used to capture high-definition images of the laser grid.
- shock waves When the wind tunnel is running, under the action of high-speed airflow, shock waves will be generated near the experimental observation area of the model in the wind tunnel experiment section, and the generation of shock waves will distort and distort the projection of the local laser grating grid on the image to form distortion image.
- the neural network model is constructed based on the recognized coordinates, the distorted corner coordinates are used as input, and the corresponding undistorted corner coordinates are used as the output true value for model training, thus constructing the neural network model.
- the distorted image collected by the camera is input into the neural network model to complete the calibration of the PIV image.
- the device and method provide a calibration solution without placing a real target, solve the problem of image distortion caused by shock waves near the model in the PIV measurement process of the high tide wind tunnel, and use the neural network model to fit the distortion, which improves the accuracy of distortion correction accuracy.
- each embodiment in this specification is described in a progressive manner, each embodiment focuses on the difference from other embodiments, and the same and similar parts of each embodiment can be referred to each other.
- the description is relatively simple, and for the related information, please refer to the description of the method part.
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Fluid Mechanics (AREA)
- Engineering & Computer Science (AREA)
- Aviation & Aerospace Engineering (AREA)
- Image Processing (AREA)
Abstract
本发明公开了一种基于激光线阵的PIV图像标定装置及方法，涉及激光测速以及图像修复技术领域，能够解决在高潮风洞中由于模型激波产生的图像畸变问题，实现畸变捕捉及修正。该装置包括：激光发射部件，用于发出具有等间距特性的激光线阵；光学组件，用于对激光线进行分光处理以在实验观测区域形成激光光栅；相机，用于获取风洞实验段模型的工况调整到PIV实验工况时的畸变后激光光栅图像；后台处理器，用于基于神经网络的畸变恢复标定算法，对所述畸变后激光光栅图像进行标定修复。本发明能够准确获取真实模型实验时的激波畸变图像，进而获得各处的图像空间分辨率。
Description
本发明涉及激光测速以及图像修复技术领域，特别是涉及一种基于激光线阵的PIV图像标定装置及方法。
粒子图像测速(PIV，Particle Image Velocimetry)技术是一种非接触流动速度场光学测量技术，该技术通过追踪由激光相机系统捕获的两帧粒子图上的粒子微团的跨帧位移，实现对粒子微团所表征的流动速度场中的运动速度的测量，并广泛应用于风洞实验中。
在记录粒子轨迹图像之前，需要对图像空间分辨率进行标定。常规方法，将黑白棋格纸的标靶放置于被测区域，通过相机采集后，利用记录的角点坐标求解畸变的数学模型，并获得各处的图像空间分辨率。但是，在高潮风洞中存在一定问题，具体为：由于风速较高，故将标靶放置于被测区域时会在局部产生激波，激波会使被测区域的局部产生光学衍射现象，从而使得记录的粒子轨迹图像产生扭曲，此处就产生较为复杂的畸变效果，进而无法准确获得各处的图像空间分辨率。
将标靶放置于被测区域，若在静止无风条件下时，就无法获得激波，无法产生畸变，进而无法获得各处的图像空间分辨率；若在有风条件下时，标靶本身就会再次引起激波，就会产生较为复杂的畸变效果，无法准确获得各处的图像空间分辨率。显然上述方法均无法获得真实模型实验时的激波畸变图像，进而无法获得各处的图像空间分辨率。
发明内容
本发明的目的是提供一种基于激光线阵的PIV图像标定装置及方法，以达到准确获取真实模型实验时的激波畸变图像的目的。
为实现上述目的，本发明提供了如下方案：
一种基于激光线阵的PIV图像标定装置，包括：
激光发射部件，用于发出具有等间距特性的激光线阵，以形成激光线阵光路；
光学组件，用于对所述激光线阵光路中的激光线进行分光处理以在实验观测区域形成激光光栅；
相机，用于获取风洞实验段模型的工况调整到PIV实验工况时的畸变后激光光栅图像；所述实验观测区域位于所述风洞实验段模型的上方；
后台处理器，用于基于神经网络的畸变恢复标定算法，对所述畸变后激光光栅图像进行标定修复，以得到重构后激光光栅图像。
可选的，所述激光发射部件包括固定架和多根激光笔；
多根所述激光笔按照平行排列方式安装在所述固定架上，且任意相邻两根所述激光笔之间的距离均相等；
在工作时，通过调整所述激光发射部件的安装角度和相邻两根所述激光笔之间的距离使得发出的激光线阵中的所有激光线平行且共面。
可选的，所述光学组件包括在所述激光线阵光路上依次布置的半反半透镜和全反射镜；
在工作时，通过调整所述半反半透镜的安装角度，使得激光线经过所述半反半透镜后，一部分所述激光线透射过去，另一部分所述激光线反射以形成第一反射激光线，然后所述第一反射激光线照射在所述实验观测区域内；
调整所述全反射镜的安装角度，使得透射过去的激光线经过所述全反射镜后被全部反射并形成第二反射激光线，然后所述第二反射激光线照射在所述实验观测区域内；
所述第一反射激光线和所述第二反射激光线在所述实验观测区域内交叉以形成交错的激光光栅。
可选的，开启高潮风洞后，所述相机用于获取风洞实验段模型的工况调整到PIV实验工况时的畸变后激光光栅图像；
开启高潮风洞前，所述相机用于获取畸变前激光光栅图像。
可选的，所述后台处理器，具体包括：
畸变二维角点坐标信息提取模块，用于采用角点检测算法，提取畸变后激光光栅图像上的激光光栅的二维角点坐标信息；
畸变前二维角点坐标信息确定模块，用于基于神经网络模型和畸变后激光光栅图像上的激光光栅的二维角点坐标信息，确定畸变前二维角点坐标信息；
重构模块，用于基于所述畸变前二维角点坐标信息，得到重构后激光光栅图像。
可选的，所述神经网络模型为三层神经网络模型；所述神经网络模型的损失函数为均方误差函数；所述神经网络模型包括两层隐含层，且每个神经元后均添加ReLU激活层；
所述神经网络模型的输入为畸变后激光光栅图像上的激光光栅的二维角点坐标信息，所述神经网络模型的输出为对应的畸变前二维角点坐标信息。
一种基于激光线阵的PIV图像标定方法，包括：
获取风洞实验段模型的工况调整到PIV实验工况时的畸变后激光光栅图像；所述激光光栅为对激光线阵光路中的激光线进行分光处理后并在实验观测区域形成的光栅；
基于神经网络的畸变恢复标定算法，对所述畸变后激光光栅图像进行标定修复，以得到重构后激光光栅图像。
可选的，还包括：
搭建激光发射部件；
所述激光发射部件包括固定架和多根激光笔，且多根所述激光笔按照平行排列方式安装在所述固定架上，任意相邻两根所述激光笔之间的距离均相等；
在工作时，通过调整所述激光发射部件的安装角度和相邻两根所述激光笔之间的距离使得发出的激光线阵中的所有激光线平行且共面。
可选的，还包括：
在激光线阵光路上依次布置半反半透镜和全反射镜；
在工作时，通过调整所述半反半透镜的安装角度，使得激光线经过所述半 反半透镜后，一部分所述激光线透射过去，另一部分所述激光线反射以形成第一反射激光线，然后所述第一反射激光线照射在实验观测区域内；
调整所述全反射镜的安装角度，使得透射过去的激光线经过所述全反射镜后被全部反射并形成第二反射激光线，然后所述第二反射激光线照射在所述实验观测区域内；
所述第一反射激光线和所述第二反射激光线在所述实验观测区域内交叉以形成交错的激光光栅。
可选的，所述基于神经网络的畸变恢复标定算法，对所述畸变后激光光栅图像进行标定修复，以得到重构后激光光栅图像，具体包括：
采用角点检测算法，提取畸变后激光光栅图像上的激光光栅的二维角点坐标信息；
基于神经网络模型和畸变后激光光栅图像上的激光光栅的二维角点坐标信息，确定畸变前二维角点坐标信息；
基于所述畸变前二维角点坐标信息，得到重构后激光光栅图像。
根据本发明提供的具体实施例，本发明公开了以下技术效果：
本发明提供了一种基于激光线阵的PIV图像标定装置及方法，本发明采集激光光栅替代传统的实物标靶，实现在存在激波的情况下进行真实模型实验标定，且不干扰真实流场。并在此基础上，采用基于神经网络的畸变恢复标定算法对畸变后激光光栅图像进行标定和畸变恢复，达到了准确获取真实模型实验时的激波畸变图像的目的，进而获得各处的图像空间分辨率。
为了更清楚地说明本发明实施例或现有技术中的技术方案，下面将对实施例中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本发明的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动性的前提下，还可以根据这些附图获得其他的附图。
图1为本发明一种基于激光线阵的PIV图像标定装置的结构示意图；
图2为本发明一种基于激光线阵的PIV图像标定方法的流程示意图；
图3为本发明一种基于激光线阵和神经网络的图像畸变标定方法的流程图；
图4为本发明神经网络的示意图；
图5为本发明实施记录的畸变前后激光光栅图；图5(a)为畸变前激光光栅图；图5(b)为畸变后激光光栅图；图5(c)为畸变修复后的激光光栅图。
下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅仅是本发明一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。
本发明的一个目的在于提供一种新的无实物标靶的PIV图像标定装置，以替代传统的实物标靶，实现在存在激波的情况下进行真实模型实验标定，且不干扰真实流场。并在此基础上，采用神经网络算法对畸变后激光光栅图像进行标定和畸变恢复。
本发明的另一个目的在于提供一种二维PIV图像标定的软硬件系统，实现在非接触无干扰的情况下获取高精度图像标定的技术方案和系统实现。
为使本发明的上述目的、特征和优点能够更加明显易懂，下面结合附图和具体实施方式对本发明作进一步详细的说明。
实施例一
本实施例提供了一种基于激光线阵的PIV图像标定装置，主要是通过激光线阵和光学组件在实验观测区域形成激光光栅作为无实物标靶。该装置应用于高马赫数风洞模型PIV实验中，该装置示意图如图1所示：
激光发射部件3，用于发出具有等间距特性的激光线阵，以形成激光线阵光路。
光学组件，用于对激光线阵光路中的激光线进行分光处理以在实验观测区域形成激光光栅。
相机6，用于获取风洞实验段模型1的工况调整到PIV实验工况时的畸变后激光光栅图像；实验观测区域位于风洞实验段模型1的上方。
后台处理器，用于基于神经网络算法，对畸变后激光光栅图像进行标定修复，以得到重构后激光光栅图像。
作为一种优选地实施方式，本实施例所述的激光线阵3包括固定架以及安装在固定架上的多根大功率的激光笔。
多根激光笔按照平行排列方式安装在固定架上，且任意相邻两根激光笔之间的距离均相等。
在工作时，通过调整激光发射部件的安装角度和相邻两根激光笔之间的距离使得发出的激光线阵中的所有激光线平行且共面。
进一步地，本实施例采用了10组连续式氦氖激光笔，其功率为1W，产生633nm波长的绿光。
作为一种优选地实施方式，本实施例所述的光学组件包括在激光线阵光路上布置的半反半透镜4和全反射镜5。此半反半透镜4和全反射镜5均位于风洞实验段模型1的上方。
在工作时，通过调整半反半透镜4的安装角度，使得激光线经过半反半透镜4后，一部分激光线透射过去，另一部分激光线反射以形成第一反射激光线，然后第一反射激光线照射在实验观测区域内。
调整全反射镜5的安装角度，使得透射过去的激光线经过全反射镜5后被全部反射并形成第二反射激光线，然后第二反射激光线照射在实验观测区域内。
第一反射激光线和第二反射激光线在实验观测区域内交叉以形成交错的激光光栅。
进一步地，本实施例所述的反半透镜4和全反射镜5的材料为镀镍等材料。
作为一种优选地实施方式，本实施例在激光光栅调整好后，需要布置PIV拍摄的相机6；调整相机6的安装位置和安装角度，确保能够捕捉到实验观测区域。
开启高潮风洞后，风洞实验段模型表面附近产生激波，此时相机6再次拍摄的激光光栅图像为畸变后激光光栅图像。故此时的相机6用于获取风洞实验段模型的工况调整到PIV实验工况时的畸变后激光光栅图像。
开启高潮风洞前，相机6用于获取畸变前激光光栅图像，以进行后续的神经网络训练。
进一步地，本实施例所述的相机6为双曝光CCD相机。通过软件控制相机6在风洞运行前后分别对激光光栅进行记录拍摄。相机数据采用高速Cameralink数据线传入后台处理器存储。
进一步地，本实施例所述的风洞用以产生均匀且稳定的高速气流，并将实验模型预先布置于实验段，实验段内布撒示踪粒子2，以增强激光光栅在图像上的显示效果。
作为一种优选地实施方式，本实施例记录畸变前后的激光光栅图像之后，采用本实施例提出的基于神经网络的畸变恢复标定算法，即可将由于激波产生的畸变后激光光栅图像进行标定修正。
首先对记录的两张激光光栅图像分别进行角点检测，提取对应的二维角点坐标信息。由于整个空间位置不发生变化，前后两张激光光栅图像检测的角点坐标一一对应。
其次构造三层神经网络结构，输入为从畸变后激光光栅图像上提取的二维角点坐标信息，输出为从畸变前激光光栅图像上提取的二维角点坐标信息。将提取的二维角点坐标信息数据划分为训练集和验证集，并采用合适的训练策略对神经网络结构进行训练直至达到拟合精度，得到需求的神经网络模型。这样的神经网络模型就隐含了整个图像的畸变信息。
最终将畸变后激光光栅图像上的每一点坐标输入到神经网络模型中，以输出畸变还原后的真实图像坐标，至此完成畸变后激光光栅图像的标定和畸变修复。
相比于传统的实物标靶方法，本实施例采用的激光光栅对真是流场无影响，同时能够记录激波引起的真实畸变图像。
实施例二
请参见图2，本实施例提供了一种基于激光线阵的PIV图像标定方法，应用于实施例一所述的一种基于激光线阵的PIV图像标定装置，该方法包括：
步骤201：获取风洞实验段模型的工况调整到PIV实验工况时的畸变后激光光栅图像；所述激光光栅为对激光线阵光路中的激光线进行分光处理后并在实验观测区域形成的光栅。
步骤202：基于神经网络的畸变恢复标定算法，对所述畸变后激光光栅图像进行标定修复，以得到重构后激光光栅图像。
进一步地，本实施例所述的一种基于激光线阵的PIV图像标定方法还包括：
搭建激光发射部件。
所述激光发射部件包括固定架和多根激光笔，且多根所述激光笔按照平行排列方式安装在所述固定架上，任意相邻两根所述激光笔之间的距离均相等；在工作时，通过调整所述激光发射部件的安装角度和相邻两根所述激光笔之间的距离使得发出的激光线阵中的所有激光线平行且共面。
进一步地，本实施例所述的一种基于激光线阵的PIV图像标定方法还包括：
在激光线阵光路上依次布置半反半透镜和全反射镜。
在工作时，通过调整所述半反半透镜的安装角度，使得激光线经过所述半反半透镜后，一部分所述激光线透射过去，另一部分所述激光线反射以形成第一反射激光线，然后所述第一反射激光线照射在实验观测区域内；调整所述全反射镜的安装角度，使得透射过去的激光线经过所述全反射镜后被全部反射并形成第二反射激光线，然后所述第二反射激光线照射在所述实验观测区域内；所述第一反射激光线和所述第二反射激光线在所述实验观测区域内交叉以形成交错的激光光栅。
进一步地，步骤202具体包括：
采用角点检测算法，提取畸变后激光光栅图像上的激光光栅的二维角点坐 标信息。
基于神经网络模型和畸变后激光光栅图像上的激光光栅的二维角点坐标信息，确定畸变前二维角点坐标信息。
基于所述畸变前二维角点坐标信息，得到重构后激光光栅图像。
实施例三
本实施例提供一种基于激光线阵和神经网络的图像畸变标定方法，请参见图3，包括：
步骤1：搭建激光线阵光路；具体为：
将多个激光笔安装在固定架上，并调整激光笔的安装间距和安装角度使激光笔发出共面平行的激光线阵，并形成激光线阵光路。
步骤2：调整光学镜片形成光栅；具体为：
步骤2.1：首先在激光线阵光路上布置半反半透镜；该半反半透镜位于风洞实验段模型的上方，该半反半透镜用于将激光线阵的一部分激光线直接透射，将激光线阵的一部分激光线反射；然后调整半反半透镜的安装位置和安装角度使得反射光线照亮实验观测区域，即流动观测区域。
步骤2.2：首先在透射光线的光路上布置全反射镜；该全反射镜位于风洞实验段模型的上方，该全反射镜用于将透射光线全部反射；其次调整全反射镜的安装位置和安装角度使得经过反射后的透射光线同样照亮实验观测区域，然后与经过半反半透镜的反射光线相交形成激光光栅。
步骤3：采集畸变前后图像；具体为：
步骤3.1：首先将相机架设到实验台架上；然后调整相机的安装位置和安装角度使相机能够准确捕捉到实验观测区域。
步骤3.2：在风洞实验段内布撒示踪粒子，以增强激光光栅的反射效果。
步骤3.3：在风洞开启前，使用相机记录激光光栅图像，此时激光光栅图像上的激光光栅未畸变。
步骤3.4：开启高潮风洞，并将风洞实验段的工况调整到PIV实验工况，此时风洞实验段模型表面产生激波，激光光栅被扭曲，最后使用相机记录畸变 后的畸变后激光光栅图像；
步骤4：角点检测；具体为：
采用角点检测算法分别提取畸变前激光光栅图像和畸变后激光光栅图像上的激光光栅的二维角点坐标信息。
角点检测算法如下：
采用设定大小的窗口在激光光栅图像的各个方向上移动，计算移动过程中窗口内灰度变化的自相关函数，如公式(1)所示：
E(u,v)＝∑
x,yw(x,y)[I(x+u,y+v)-I(x,y)]
2 (1)；
其中，(u,v)为窗口的大小；w为窗口的权重，取为1；I为图像像素灰度值；(x，y)为像素坐标。
经过泰勒展开后可以将自相关函数E写为：
M的计算公式为：
定义角点相应函数R为：
R＝detM-k(traceM)
2 (4)；
traceM＝λ
1+λ
2 (5)；
detM＝λ
1λ
2 (6)；
其中，traceM为矩阵M的迹；detM为矩阵M的秩；λ
1和λ
2为矩阵M的特征值；k为经验常数，一般取0.04～0.06。
当检测到R大于0时即可定位角点位置并提取二维角点坐标信息。
步骤5：构建神经网络。
请参见图4，神经网络包含两层隐含层，每个神经元后添加ReLU激活层。神经元的计算数学表达式如下所示：
采用均方误差函数计算神经网络的损失，即神经网络的损失函数为：
其中，(x,y)为神经网络输出的预测坐标，(x',y')为神经网络输出的真实坐标，即通过角点检测算法确定出的畸变前二维角点坐标。n为训练过程中一个批次的batch个数。
步骤6：训练，以得到神经网络模型；此神经网络模型的输入为畸变后的二维角点坐标信息，输出为畸变前的二维角点坐标信息。
将之前检测到的畸变前的二维角点坐标信息和畸变后的二维角点坐标信息组成训练集样本，并以8个样本为一组作为一个batch进行神经网络权重的训练。
其中，选用Adam优化学习算法进行神经网络优化训练。优化流程如下:
步骤6.1：初始化神经网络中的权重记为θ
0，初始化一阶动量矩m
0、二阶动量矩v
0，学习率α＝0.00001、参数权重β
1＝0.9、β
2＝0.999、ε＝10
-8。
步骤6.2：在当前权重下，带入一个批次样本计算神经网络输出损失Loss
t(θ
t-1)，并求其对权重θ的梯度，其计算公式如下：
其中，t表示当前迭代次数，t-1为上一次迭代次数。
步骤6.:3：计算有偏一阶动量矩和有偏二阶动量矩，计算公式如下：
m
t＝β
1×m
t-1+(1-β
1)g
t (10)；
v
t＝β
2×v
t-1+(1-β
2)g
t
2 (11)；
步骤6.4：计算无偏一阶动量矩和无偏二阶动量矩，计算公式如下：
步骤6.5：计算并更新神经网络权重，计算公式如下：
步骤6.6：重复步骤6.2至步骤6.5，直至损失不再下降即可认为训练完成。
步骤6.7：采用未使用的二维角点坐标信息作为验证集验证神经网络模型预测的准确性，当准确度高于90％以上时即认为神经网络模型符合要求。
步骤7：标定完成，图像畸变修复。
获得包含畸变信息的神经网络模型后，将畸变后激光光栅图像上的每个像素点(x,y,I)提取出来；其中，I为像素上灰度信息。将坐标(x,y)输入到包含畸变信息的神经网络模型后，预测得到畸变前激光光栅图像上的坐标(x',y')，遍历畸变后激光光栅图像上所有的像素点(x,y,I)，即可重构出对应的畸变前激光光栅图像，其效果请参见图5。
本发明公开了一种基于激光线阵的PIV图像标定装置及方法，包括：发出一系列等间距的激光线阵的激光发射部件，在激光线阵光路上布置多组光学镜片以对激光线进行反射和投射，即实现激光线阵的分光，并通过合理布置光学镜片的安装位置和安装角度以在风洞实验段模型附近形成激光线阵垂直交叉的激光栅格。在风洞实验段模型的观测窗口外布置相机，并用相机捕获激光栅格的高清图像。
当风洞运行时，在高速气流的作用下，风洞实验段模型的实验观测区域附近将产生激波，激波的产生使得局部激光光栅网格在图像上的投影产生扭曲畸变，以形成畸变图像。
记录风洞运行前后两帧激光栅格的高清图像，并采用角点检测算法对风洞运行前后的记录的两张图像分别进行检测，提取所有识别到的激光光栅栅格的交点的坐标，并基于识别的坐标构建神经网络模型，将畸变后的角点坐标作为输入，对应的未畸变角点坐标作为输出真值进行模型训练，这样就构造了神经网络模型。将相机采集的畸变图像输入到神经网络模型中完成了PIV图像的标定。
该装置和方法提供了无需放置标靶实物的标定解决方案，解决了高潮风洞 PIV测量过程中由于模型附近激波引起的图像畸变问题，并采用神经网络模型拟合畸变，提高了畸变修正的准确性。
本说明书中各个实施例采用递进的方式描述，每个实施例重点说明的都是与其他实施例的不同之处，各个实施例之间相同相似部分互相参见即可。对于实施例公开的系统而言，由于其与实施例公开的方法相对应，所以描述的比较简单，相关之处参见方法部分说明即可。
本文中应用了具体个例对本发明的原理及实施方式进行了阐述，以上实施例的说明只是用于帮助理解本发明的方法及其核心思想；同时，对于本领域的一般技术人员，依据本发明的思想，在具体实施方式及应用范围上均会有改变之处。综上所述，本说明书内容不应理解为对本发明的限制。
Claims (10)
- 一种基于激光线阵的PIV图像标定装置，其特征在于，包括：激光发射部件，用于发出具有等间距特性的激光线阵，以形成激光线阵光路；光学组件，用于对所述激光线阵光路中的激光线进行分光处理以在实验观测区域形成激光光栅；相机，用于获取风洞实验段模型的工况调整到PIV实验工况时的畸变后激光光栅图像；所述实验观测区域位于所述风洞实验段模型的上方；后台处理器，用于基于神经网络的畸变恢复标定算法，对所述畸变后激光光栅图像进行标定修复，以得到重构后激光光栅图像。
- 根据权利要求1所述的一种基于激光线阵的PIV图像标定装置，其特征在于，所述激光发射部件包括固定架和多根激光笔；多根所述激光笔按照平行排列方式安装在所述固定架上，且任意相邻两根所述激光笔之间的距离均相等；在工作时，通过调整所述激光发射部件的安装角度和相邻两根所述激光笔之间的距离使得发出的激光线阵中的所有激光线平行且共面。
- 根据权利要求1所述的一种基于激光线阵的PIV图像标定装置，其特征在于，所述光学组件包括在所述激光线阵光路上依次布置的半反半透镜和全反射镜；在工作时，通过调整所述半反半透镜的安装角度，使得激光线经过所述半反半透镜后，一部分所述激光线透射过去，另一部分所述激光线反射以形成第一反射激光线，然后所述第一反射激光线照射在所述实验观测区域内；调整所述全反射镜的安装角度，使得透射过去的激光线经过所述全反射镜后被全部反射并形成第二反射激光线，然后所述第二反射激光线照射在所述实验观测区域内；所述第一反射激光线和所述第二反射激光线在所述实验观测区域内交叉以形成交错的激光光栅。
- 根据权利要求1所述的一种基于激光线阵的PIV图像标定装置，其特征在于，开启高潮风洞后，所述相机用于获取风洞实验段模型的工况调整到PIV实验工况时的畸变后激光光栅图像；开启高潮风洞前，所述相机用于获取畸变前激光光栅图像。
- 根据权利要求1所述的一种基于激光线阵的PIV图像标定装置，其特征在于，所述后台处理器，具体包括：畸变二维角点坐标信息提取模块，用于采用角点检测算法，提取畸变后激光光栅图像上的激光光栅的二维角点坐标信息；畸变前二维角点坐标信息确定模块，用于基于神经网络模型和畸变后激光光栅图像上的激光光栅的二维角点坐标信息，确定畸变前二维角点坐标信息；重构模块，用于基于所述畸变前二维角点坐标信息，得到重构后激光光栅图像。
- 根据权利要求5所述的一种基于激光线阵的PIV图像标定装置，其特征在于，所述神经网络模型为三层神经网络模型；所述神经网络模型的损失函数为均方误差函数；所述神经网络模型包括两层隐含层，且每个神经元后均添加ReLU激活层；所述神经网络模型的输入为畸变后激光光栅图像上的激光光栅的二维角点坐标信息，所述神经网络模型的输出为对应的畸变前二维角点坐标信息。
- 一种应用于权利要求1所述的基于激光线阵的PIV图像标定装置的标定方法，其特征在于，包括：获取风洞实验段模型的工况调整到PIV实验工况时的畸变后激光光栅图像；所述激光光栅为对激光线阵光路中的激光线进行分光处理后并在实验观测区域形成的光栅；基于神经网络的畸变恢复标定算法，对所述畸变后激光光栅图像进行标定修复，以得到重构后激光光栅图像。
- 根据权利要求7所述的标定方法，其特征在于，还包括：搭建激光发射部件；所述激光发射部件包括固定架和多根激光笔，且多根所述激光笔按照平行排列方式安装在所述固定架上，任意相邻两根所述激光笔之间的距离均相等；在工作时，通过调整所述激光发射部件的安装角度和相邻两根所述激光笔之间的距离使得发出的激光线阵中的所有激光线平行且共面。
- 根据权利要求7所述的标定方法，其特征在于，还包括：在激光线阵光路上依次布置半反半透镜和全反射镜；在工作时，通过调整所述半反半透镜的安装角度，使得激光线经过所述半反半透镜后，一部分所述激光线透射过去，另一部分所述激光线反射以形成第一反射激光线，然后所述第一反射激光线照射在实验观测区域内；调整所述全反射镜的安装角度，使得透射过去的激光线经过所述全反射镜后被全部反射并形成第二反射激光线，然后所述第二反射激光线照射在所述实验观测区域内；所述第一反射激光线和所述第二反射激光线在所述实验观测区域内交叉以形成交错的激光光栅。
- 根据权利要求7所述的标定方法，其特征在于，所述基于神经网络的畸变恢复标定算法，对所述畸变后激光光栅图像进行标定修复，以得到重构后激光光栅图像，具体包括：采用角点检测算法，提取畸变后激光光栅图像上的激光光栅的二维角点坐标信息；基于神经网络模型和畸变后激光光栅图像上的激光光栅的二维角点坐标信息，确定畸变前二维角点坐标信息；基于所述畸变前二维角点坐标信息，得到重构后激光光栅图像。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| PCT/CN2021/114416 WO2023023961A1 (zh) | 2021-08-25 | 2021-08-25 | 一种基于激光线阵的piv图像标定装置及方法 |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| PCT/CN2021/114416 WO2023023961A1 (zh) | 2021-08-25 | 2021-08-25 | 一种基于激光线阵的piv图像标定装置及方法 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2023023961A1 true WO2023023961A1 (zh) | 2023-03-02 |
Family
ID=85321600
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/CN2021/114416 WO2023023961A1 (zh) | 2021-08-25 | 2021-08-25 | 一种基于激光线阵的piv图像标定装置及方法 |
Country Status (1)
| Country | Link |
|---|---|
| WO (1) | WO2023023961A1 (zh) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117805434A (zh) * | 2024-03-01 | 2024-04-02 | 中国空气动力研究与发展中心低速空气动力研究所 | 用于时空演化壁面湍流边界层的spiv测量、标定装置及方法 |
| CN117805434B (zh) * | 2024-03-01 | 2024-06-04 | 中国空气动力研究与发展中心低速空气动力研究所 | 用于时空演化壁面湍流边界层的spiv测量、标定装置及方法 |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130242286A1 (en) * | 2012-03-16 | 2013-09-19 | Honda Motor Co., Ltd. | Particle image velocimetry system for three-dimensional space |
| CN105488807A (zh) * | 2015-12-25 | 2016-04-13 | 傅之成 | 一种远心镜头的标定和矫正方法 |
| CN111347431A (zh) * | 2020-04-16 | 2020-06-30 | 广东工业大学 | 一种手持工具示教的机器人示教喷涂方法和装置 |
| CN112577697A (zh) * | 2020-12-29 | 2021-03-30 | 中国航天空气动力技术研究院 | 一种超声速风洞流场多相机三维标定装置及方法 |
-
2021
- 2021-08-25 WO PCT/CN2021/114416 patent/WO2023023961A1/zh unknown
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130242286A1 (en) * | 2012-03-16 | 2013-09-19 | Honda Motor Co., Ltd. | Particle image velocimetry system for three-dimensional space |
| CN105488807A (zh) * | 2015-12-25 | 2016-04-13 | 傅之成 | 一种远心镜头的标定和矫正方法 |
| CN111347431A (zh) * | 2020-04-16 | 2020-06-30 | 广东工业大学 | 一种手持工具示教的机器人示教喷涂方法和装置 |
| CN112577697A (zh) * | 2020-12-29 | 2021-03-30 | 中国航天空气动力技术研究院 | 一种超声速风洞流场多相机三维标定装置及方法 |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117805434A (zh) * | 2024-03-01 | 2024-04-02 | 中国空气动力研究与发展中心低速空气动力研究所 | 用于时空演化壁面湍流边界层的spiv测量、标定装置及方法 |
| CN117805434B (zh) * | 2024-03-01 | 2024-06-04 | 中国空气动力研究与发展中心低速空气动力研究所 | 用于时空演化壁面湍流边界层的spiv测量、标定装置及方法 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Shi et al. | Parametric study on light field volumetric particle image velocimetry | |
| Besnard et al. | “Finite-element” displacement fields analysis from digital images: application to Portevin–Le Châtelier bands | |
| CN102183214B (zh) | 一种大口径非球面镜结构光检测方法 | |
| US7577309B2 (en) | Direct vision sensor for 3D computer vision, digital imaging, and digital video | |
| CN111476827B (zh) | 目标跟踪方法、系统、电子装置及存储介质 | |
| Yang et al. | Kernelized Lagrangian particle tracking | |
| CN102239426A (zh) | 图像数据的提供 | |
| CN106290256A (zh) | 基于视频测量的定量背景纹影方法 | |
| US20230025527A1 (en) | Quantitative analysis method and system for attention based on line-of-sight estimation neural network | |
| CN115100294A (zh) | 基于直线特征的事件相机标定方法、装置及设备 | |
| CN113687099B (zh) | 一种基于激光线阵的piv图像标定装置及方法 | |
| CN117928875A (zh) | 一种风洞流场的时间分辨偏振成像装置及方法 | |
| Raffel et al. | Techniques for 3D-PIV | |
| CN113298700A (zh) | 一种在散射场景中的高分辨图像重构方法 | |
| WO2023023961A1 (zh) | 一种基于激光线阵的piv图像标定装置及方法 | |
| Zhu et al. | Full-field modal identification using reliability-guided frequency-domain-based digital image correlation method based on multi-camera system | |
| CN110108271B (zh) | 一种气动光学效应引起的星光偏折补偿方法 | |
| Loktev et al. | Image Blur Simulation for the Estimation of the Behavior of Real Objects by Monitoring Systems. | |
| CN114119747B (zh) | 一种基于pmd波前检测的三维流场流动显示方法 | |
| KR101237128B1 (ko) | 토모그래픽 피티브이를 이용한 유동장 해석방법 | |
| Arredondo et al. | A methodology for the systematic assessment of underwater video processing algorithms | |
| Yaryshev et al. | Development of a digital camera-based method for bridge deformation measurement | |
| CN114065650A (zh) | 基于深度学习的材料裂纹尖端多尺度应变场测量跟踪方法 | |
| Zhao et al. | Nodes2STRNet for structural dense displacement recognition by deformable mesh model and motion representation | |
| CN113295701A (zh) | 基于数字图像对剪力墙损伤的测量方法、设备和介质 |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 21954502 Country of ref document: EP Kind code of ref document: A1 |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |