A high instantaneous speed model attitude angle on-line measuring method

Abstract

The application discloses a high instantaneous speed model attitude angle online measurement method, relates to the field of aerodynamic test technology, and comprises the following steps: n cooperative marker points are arranged on the surface of a measured model, any three cooperative marker points are not collinear, each cooperative marker point emits or reflects one of red, green and blue light, and the light emitted or reflected by two adjacent cooperative marker points is of different colors; three one-dimensional phase mechanisms are used to form a three-dimensional coordinate measuring device, and the three-dimensional coordinate measuring device is used for synchronously measuring the three-dimensional coordinates of the n cooperative marker points; when n is 3, the attitude angle of the measured model is obtained according to the three-dimensional coordinates of the n cooperative marker points; when n is greater than 3, a transformation matrix of a moment t relative to a moment t0 is calculated by using a least square method, and the attitude angle of the measured model is calculated based on the transformation matrix and the three-dimensional coordinates of the n cooperative marker points, so that the problem that displacement deviation between cooperative marker points at different moments influences the attitude calculation precision can be effectively solved.

Description

Technical Field

[0001] This invention relates to the field of aerodynamic testing technology, specifically to an image processing method, system, device, and medium, and a counting method and system. Background Technology

[0002] During the development of aircraft, ground simulation tests need to be conducted in wind tunnels to test aerodynamic parameters and other parameters under different flight attitudes.

[0003] Currently, methods used for measuring the attitude angle of wind tunnel test models mainly include: accelerometer-based attitude measurement, angle of attack measurement based on model support mechanisms, Optotrak optical measurement, and photogrammetry. Among these, accelerometer-based attitude measurement has good static response but is easily affected by vibration. Photogrammetry methods, such as... Figure 1 As shown, artificial markers are pasted on the model surface, and the three-dimensional coordinates of these markers are obtained using the principle of binocular vision measurement, thereby calculating the model's attitude angles. Due to limitations in the frame rate of imaging devices (high-resolution cameras typically have frame rates below 100fps) and image processing speed, photogrammetry methods struggle to meet the demands of online measurement tasks exceeding 1K. Even with high-speed cameras, which can achieve image sampling frequencies above 10K, online measurement is impossible due to the massive amount of image data that needs processing. Furthermore, high-speed cameras are expensive (generally exceeding 500,000 RMB per unit), and the large volume of captured image data prevents prolonged shooting and measurement. Optotrak (… Figure 2 Multiple LED beads are fabricated on the surface of the model. A controller sequentially illuminates the beads, and three line-scan cameras simultaneously capture images of the illuminated beads. The three-dimensional coordinates of the beads are calculated based on the measurement planes formed by the three line-scan cameras in pairs. Figure 3 By sequentially illuminating all the LEDs in this manner, several measurement points can be acquired for calculating the model's attitude angles. Compared to photogrammetry, Optotrak uses a linear array image sensor, offering high imaging resolution (up to 4k pixels or more), high sampling frequency (up to 20kHz or more), and the small data size of the linear array image allows for online processing, thus obtaining high-precision, online measurement results. However, as... Figure 4 As shown, when the measured object has a high instantaneous velocity, the three-dimensional coordinates measured by the sequence method differ from the true coordinates. Figure 4 In the diagram, the solid circle represents the positions of the three cooperative markers during the first measurement, namely 1-1, 1-2, and 1-3. The dashed circle represents the actual measured positions of the second and third cooperative markers, which have offsets of d1 and d2 from their positions during the first measurement. These offsets will cause errors in the attitude calculation.

[0004] Assuming the measured object's speed is 100 m / s and the linear array camera's sampling frequency is 10 kHz, the displacement of the LED during the sequence measurement is 10 mm. Taking angle of attack measurement as an example, assuming the distance between two measuring points is 100 mm, a 10 mm displacement can introduce a 10% angle calculation error.

[0005] To address this issue, the existing technology, "Research on Attitude Measurement of Spatial Objects Based on Multi-Linear CCD with Multi-Point Cooperative Targets," addresses the problem by rationally arranging the positions of three cooperative marker points. Figure 5 This technique enables online measurement of model attitude angles within the range of -10° to +10°. However, the method has the following drawbacks:

[0006] 1) The number of cooperative markers is too small. Three cooperative markers are the minimum requirement for attitude angle measurement. When there is noise pollution in the measurement results, it is difficult to guarantee the attitude measurement accuracy when the attitude angle is directly calculated from three cooperative markers.

[0007] 2) Within the framework of this technology, if the number of marker points is increased and methods such as least squares estimation are used to improve the accuracy of attitude angle estimation, then multiple marker points need to be identified and matched. Since each marker point only represents a peak or valley in one-dimensional imaging, there are no distinguishing features, and the identification and matching of marker points before and after model movement is not utilized.

[0008] 3) such as Figure 6 As shown, when any two of the three cooperative markers are collinear, it becomes impossible to separate the two cooperative markers from the linear array image, thus preventing the acquisition of the three-dimensional coordinates of the three cooperative markers and the measurement of attitude angles. Therefore, the method in this paper can only achieve angle measurements within the range of -10° to +10° and cannot be used for larger angle measurements. In actual wind tunnel tests, especially in high angle-of-attack and model free-flight tests, the angle often exceeds 10 degrees, so this method also cannot meet the needs of practical applications. When the number of cooperative markers is larger, the probability of two markers collinearly occluding is greater, posing a greater challenge to marker identification and front-back correlation matching. Summary of the Invention

[0009] Therefore, in wind tunnel tests, when the model is in virtual flight, high-speed flow field, or vibrating at high speed, and the model has high instantaneous velocity, the existing Optotrak measurement method cannot accurately measure the attitude, and the photogrammetry method cannot perform online measurement. The existing technology "Research on Attitude Measurement of Spatial Objects Based on Multi-Linear CCD Based on Multi-Point Cooperative Targets" has the problems of small adaptability angle range and attitude angle measurement accuracy being easily affected by the calculation error of the coordinates of the three cooperative markers. This invention provides an online measurement method for the attitude angle of models with high instantaneous velocity.

[0010] To achieve the above-mentioned objective, this invention provides a method for online measurement of attitude angles of a high instantaneous velocity model, the method comprising:

[0011] n cooperative marker points are set on the surface of the model to be tested, where n is an integer greater than or equal to 3. Any three cooperative marker points are not collinear. Each cooperative marker point emits or reflects one of the three colors of light: red, green, and blue. The colors of light emitted or reflected by two adjacent cooperative marker points are different.

[0012] A three-dimensional coordinate measuring device is constructed using three one-dimensional cameras, and the three-dimensional coordinates of n cooperative marker points are measured simultaneously using the three-dimensional coordinate measuring device.

[0013] When n=3, the attitude angle of the model under test is obtained by solving the three-dimensional coordinates of n cooperative marker points;

[0014] When n>3, the least squares method is used to calculate the transformation matrix of time t relative to time t0, and the attitude angle of the model under test is calculated based on the transformation matrix and the three-dimensional coordinates of n cooperative marker points.

[0015] In particular, for a rigid body object under test, once its three-dimensional spatial coordinates are known through the above method, its posture can be calculated.

[0016] In some embodiments, when n>3, any two cooperative markers emit or reflect light of different colors. When two points are collinear, they can be separated and identified by their spectra. If they are the same color, they are difficult to distinguish.

[0017] In some embodiments, the one-dimensional camera includes a cylindrical mirror and an RGB three-line array image sensor. The cylindrical mirror is used to focus the light rays of the cooperative marker points on the surface of the model under test onto the RGB three-line array image sensor. The RGB three-line array image sensor includes three linear array imaging units for sensing red, green, and blue light, respectively. The three linear array imaging units are aligned end to end and arranged in parallel. This arrangement is to reduce measurement errors because when the same marker point is imaged in three channels, we assume that they are collinear. However, there is a real distance between them, and a large distance will affect the measurement accuracy. Therefore, arranging them side by side in parallel is optimal.

[0018] Wherein, in some embodiments, the light reflected or emitted by the cooperative marker appears as a circular light spot. Before the light of the marker propagates to the linear image sensor, it passes through a cylindrical lens that flattens the light spot. When performing measurement, the sub-pixel center coordinates of the technical marker on the linear array image are required. If the marker is circular, after being compressed by the cylindrical lens, it will present an intensity curve with Gaussian distribution, and the center of this Gaussian distribution is the center of the marker. This is a property unique to circular markers. If the marker is not circular, it is difficult to accurately obtain the center position of the marker on linear array imaging.

[0019] Wherein, in some embodiments, the cooperative marker is fabricated with a light-emitting diode or an LED bead; or, the cooperative marker is fabricated as a disk pattern with one of red, green and blue paints, adhered to the surface of the measured model and illuminated with white light.

[0020] Wherein, in some embodiments, the R, G and B linear array imaging units in the RGB three-linear array image sensor respectively image the cooperative marker to obtain Ir, Ig and Ib linear array images; the central sub-pixel coordinates of the cooperative marker are extracted and obtained from the Ir, Ig and Ib linear array images, and the three-dimensional coordinates of the cooperative marker are calculated based on the central sub-pixel coordinates of the cooperative marker, the imaging model of one-dimensional cameras and the geometric positional relationship of the three one-dimensional cameras.

[0021] Wherein, in some embodiments, the specific method for extracting and obtaining the central sub-pixel coordinate Pr of the cooperative marker from the Ir linear array image is as follows:

[0022] Step a-1: set a maximum brightness pixel marker vector Vr, the length of Vr is the same as that of Ir, and the initial values of all elements in Vr are set to 0;

[0023] Step a-2: according to the marker vector Vr, find the pixel Rm with the maximum brightness among all pixels marked as 0 in Vr in the Ir linear array image, and set the element at the position of Rm in Vr to 1;

[0024] Step a-3: take out the pixels at the position of Rm from the Ig and Ib linear array images to obtain Ig(Rm) and Ib(Rm); when Ig(Rm)<T and Ib(Rm)<T, in the Ir linear array image, take the pixel values in the local neighborhood of 2*s+1 pixels with Rm as the center, calculate and obtain the central sub-pixel coordinate Pr, wherein T is a preset value and s is the window size;

[0025] when Ig(Rm)>T and Ib(Rm)>T, go back to step a-2.

[0026] Wherein, in some embodiments, the specific method for extracting and obtaining the central sub-pixel coordinate Pg of the cooperative marker from the Ig linear array image is as follows:

[0027] Step b-1: set a maximum brightness pixel marking vector Vg, the length of Vg is the same as that of Ig, and all elements in Vg are initially set to 0;

[0028] Step b-2: according to the marking vector Vg, find the pixel Gm with maximum brightness among all pixels marked 0 in Vg in the Ig linear array image, and set the element at the position of Gm in Vg to 1;

[0029] Step b-3: extract pixels at position Gm from Ir and Ib linear array images to obtain Ir(Gm) and Ib(Gm); when Ir(Gm)<T and Ib(Gm)<T, in the Ig linear array image, with Gm as the center, take pixel values in the local neighborhood of 2*s+1 pixels, and calculate to obtain the center sub-pixel coordinate Pg of the cooperative marking point;

[0030] when Ir(Gm)>T and Ib(Gm)>T, go back to step b-2.

[0031] wherein, in some embodiments, the specific method for extracting and obtaining the center sub-pixel coordinate Pb of the cooperative marking point from the Ib linear array image is:

[0032] Step c-1: set a maximum brightness pixel marking vector Vb, the length of Vb is the same as that of Ib, and all elements in Vb are initially set to 0;

[0033] Step c-2: according to the marking vector Vb, find the pixel Bm with maximum brightness among all pixels marked 0 in Vb in the Ib linear array image, and set the element at the position of Bm in Vb to 1;

[0034] Step c-3: extract pixels at position Bm from Ir and Ig linear array images to obtain Ir(Bm) and Ig(Bm); when Ir(Bm)<T and Ig(Bm)<T, in the Ib linear array image, with Bm as the center, take pixel values in the local neighborhood of 2*s+1 pixels, and calculate the center sub-pixel coordinate Pb of the cooperative marking point;

[0035] when Ir(Bm)>T and Ig(Bm)>T, go back to step c-2.

[0036] wherein, the channel data with the maximum intensity among the three channels can be found through the above method. The above method is an effective marking point detection and identification method under multi-channel imaging conditions.

[0037] One or more technical solutions provided by the present invention have at least the following technical effects or advantages:

[0038] Compared with the existing model pose measurement technology, especially compared with the Optotrak method, as Figure 6As shown, this invention employs an RGB three-line array image sensor to perceive red, green, and blue cooperative marker points on the model surface, effectively solving the occlusion problem of collinear cooperative marker points in the existing method, "Research on Attitude Measurement of Spatial Objects Based on Multi-Linear CCD Based on Multi-Point Cooperative Targets." When marker points are collinear, we can distinguish between collinear (overlapping) marker points through the spectrum of different colors in multi-channel data, thus enabling effective measurement of the marker points. Furthermore, it simultaneously acquires the three-dimensional coordinates of three non-collinear cooperative marker points, effectively avoiding the problem of displacement deviations between cooperative marker points at different times affecting the attitude calculation accuracy, which exists in existing asynchronous measurement methods. Moreover, compared to photogrammetry, this invention only requires processing three linear array images, compared to two area array images required by photogrammetry. This results in smaller data volume, higher sampling rate, and enables rapid online attitude measurement. It has significant application value for improving the high accuracy and rapid online measurement of attitude angles of wind tunnel test models. Attached Figure Description

[0039] The accompanying drawings, which are provided to further illustrate embodiments of the invention and constitute a part of this invention, are not intended to limit the scope of the invention.

[0040] Figure 1 Schematic diagram of photogrammetry principles;

[0041] Figure 2 This is a schematic diagram of the Optotrak measurement system.

[0042] Figure 3 A schematic diagram illustrating the principle of three-dimensional coordinate measurement using three one-dimensional cameras;

[0043] Figure 4 A schematic diagram illustrating the displacement deviation between measuring points introduced by the sequence measurement method;

[0044] Figure 5 A schematic diagram of an existing three-cooperative marker model attitude measurement system;

[0045] Figure 6 A schematic diagram of the occlusion problem in the pose measurement of an existing three-cooperative marker model;

[0046] Figure 7 A schematic diagram illustrating how the three-color linear array imaging of this invention solves the occlusion problem;

[0047] Figure 8 Schematic diagram of the apparatus for implementing the method of the present invention;

[0048] Figure 9 Schematic diagram of an RGB three-color one-dimensional camera;

[0049] In the diagram, 1 is a cooperative marker point, 2 is a single-channel linear array image sensor, 3 is a red marker point, 4 is a green marker point, 5 is a blue marker point, 6 is an RGB three-line array image sensor, 7 is a blue linear array imaging unit, 8 is a green linear array imaging unit, 9 is a red linear array imaging unit, 10 is the model under test, 11-1, 11-2, and 11-3 are all RGB three-color one-dimensional cameras, 12 is a cylindrical lens, 13 is an image ray, 14 to 16 are three linear array CCDs respectively, 17 to 19 are three cylindrical lenses respectively, 20 is the point under test, 21 is a computer, 22 is a camera, 23 is a lamp, 24 is a model, 25 is a marker point, and 26 is the object under test. Detailed Implementation

[0050] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, where there is no conflict, the embodiments of the present invention and the features thereof can be combined with each other.

[0051] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and therefore the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0052] Example 1

[0053] To address the problem that existing model attitude measurement methods cannot be used for online measurement of attitude angles of high instantaneous velocity models, this invention provides an online attitude measurement method for high instantaneous velocity models.

[0054] The technical solution provided by this invention is as follows: n non-collinear cooperative marker points are set on the surface of the model under test. The n cooperative marker points reflect or emit red, green and blue light respectively, and the value of n is greater than or equal to 3. A three-dimensional coordinate measuring device is constructed by three red, green and blue one-dimensional cameras to simultaneously measure the three-dimensional coordinates of the n cooperative marker points. The attitude angle of the model under test is calculated based on the three-dimensional coordinates of the n cooperative marker points.

[0055] When n=3, the method in the existing technology "Research on Attitude Measurement of Spatial Objects Based on Multi-Linear CCD Based on Multi-Point Cooperative Targets" is adopted. The embodiments of this invention will not elaborate on the method.

[0056] When n>3, when setting cooperative markers, any two adjacent markers are made to have different colors. The least squares method is used to estimate the transformation matrix of time t relative to time t0, which is used to accurately estimate the model attitude angles and reduce the sensitivity of model attitude angle estimation to the accuracy of the 3D coordinate calculation of individual cooperative markers. The least squares method is a mathematical tool widely used in many disciplines such as error estimation, uncertainty, system identification and prediction, and forecasting. This embodiment of the invention will not elaborate on this further.

[0057] In this embodiment, the tested model is assumed to be a rigid body. Three non-collinear points define a plane. Once the coordinates of these three points are known, the attitude change can be directly calculated using coordinate transformation formulas. When there are more than three points, least squares estimation is used. The attitude angle of the tested model is calculated based on the transformation matrix and the 3D coordinates of n cooperative marker points. If the positions of the n marker points at time t0 and the transformation matrix from time t to t0 are known, then the attitude of the tested model relative to time t0 is the 3D coordinates at time t0 multiplied by the transformation matrix. The method for obtaining the transformation matrix from time t to t0 can be found in the Fundamentals of Robot Kinematics - Rigid Body Attitude Description, which will not be elaborated upon in this embodiment.

[0058] Further: When n>3, n cooperative markers with different colors can be set, and each marker has a different color.

[0059] The red, green, and blue three-color one-dimensional camera consists of a cylindrical mirror and an RGB three-line array image sensor. The cylindrical mirror is used to focus the light from the three-color cooperative marker points on the model surface onto the RGB three-line array image sensor. The RGB three-line array image sensor contains three linear array imaging units that respectively sense red, green, and blue light. The three linear array imaging units are aligned end to end and arranged in parallel.

[0060] A one-dimensional camera using red, green, and blue colors is simply referred to as an RGB three-color one-dimensional camera.

[0061] The light reflected or emitted by the cooperative marker points appears as circular light spots.

[0062] The cooperative markers are made of light-emitting diodes or LED beads, emitting circular red, blue, and green light respectively.

[0063] The cooperative markers can also be made into disc patterns using red, green, and blue paints, adhered to the model surface, and illuminated with white light.

[0064] The three R, G and B linear array imaging units in the RGB three-line array image sensor image the three red, green and blue cooperative marker points simultaneously, and obtain Ir, Ig and Ib linear array images respectively; the sub-pixel coordinates of the centers of the three cooperative marker points are extracted from the Ir, Ig and Ib linear array images, and then the three-dimensional coordinates of the three cooperative marker points are calculated according to the imaging model of one-dimensional cameras and the geometric positional relationship of the three one-dimensional cameras (for details, refer to the prior art *Research on Pose Measurement of Multi-linear Array CCD Space Objects Based on Multi-point Cooperative Targets*).

[0065] The method for extracting the center sub-pixel coordinates of three separate R, G and B cooperative marker points from the Ir, Ig and Ib linear array images is as follows:

[0066] a) Extract the center sub-pixel coordinate Pr of the red marker point:

[0067] Step a-1: Set a maximum brightness pixel marking vector Vr, the length of Vr is the same as that of Ir, and the initial values of all elements in Vr are set to 0;

[0068] Step a-2: According to the marking vector Vr, find the pixel Rm with the maximum brightness among all pixels marked 0 in Vr in the Ir linear array image, and then set the element at the Rm position in Vr to 1;

[0069] Step a-3: Take out the pixels at the Rm position from the Ig and Ib linear array images: Ig(Rm) and Ib(Rm); when Ig(Rm)<t and Ib(Rm)<t, in the Ir linear array image, take the pixel values in the local neighborhood of 2*s+1 pixels centered on Rm, and calculate the sub-pixel center coordinate Pr of the red marker point by Gaussian fitting or centroid method;

[0070] When Ig(Rm)>t and Ib(Rm)>t, go to step a-2;

[0071] b) Extract the center sub-pixel coordinate Pg of the green marker point:

[0072] Step b-1: Set a maximum brightness pixel marking vector Vg, the length of Vg is the same as that of Ig, and the initial values of all elements in Vg are set to 0;

[0073] Step b-2: According to the marking vector Vg, find the pixel Gm with the maximum brightness among all pixels marked 0 in Vg in the Ig linear array image, and set the element at the Gm position in Vg to 1;

[0074] Step b-3: Take out the pixels at the Gm position from the Ir and Ib linear array images: Ir(Gm) and Ib(Gm);

[0075] When Ir(Gm)<t and Ib(Gm)<t, in the Ig linear array image, with Gm as the center, the pixel values in the local neighborhood of 2*s+1 pixels are taken, and Gaussian fitting or centroid method is used to calculate the sub-pixel center coordinate Pg of the blue marking point;

[0076] When Ir(Gm)>t and Ib(Gm)>t, go to step b-2;

[0077] c) Extract the sub-pixel coordinate Pb of the center of the blue marking point:

[0078] Step c-1: Set the maximum brightness pixel marking vector Vb, the length of Vb is the same as Ib, and the initial values of all elements in Vb are set to 0;

[0079] Step c-2: According to the marking vector Vb, find the maximum brightness pixel Bm among all pixels marked as 0 in Vb in the Ib linear array image, and set the element at the Bm position in Vb to 1;

[0080] Take out the pixels at the Bm position from the Ir and Ig linear array images: Ir(Bm), Ig(Bm);

[0081] When Ir(Bm)<t and Ig(Bm)<t, in the Ib linear array image, with Bm as the center, the pixel values in the local neighborhood of 2*s+1 pixels are taken, and Gaussian fitting or centroid method is used to calculate the sub-pixel center coordinate Pb of the blue marking point;

[0082] When Ir(Bm)>t and Ig(Bm)>t, go to step c-2;

[0083] Wherein, the value range of t is 1 to 10000, and the value range of s is 1 to 100.

[0084] Wherein, Gaussian fitting or centroid method are commonly used calculation methods in this field, and will not be described correspondingly in the embodiments of the present invention.

[0085] Embodiment 2:

[0086] On the basis of Embodiment 1, three kinds of circular marking points are made by using red, green and blue pigments: red marking point 3, green marking point 4, and blue marking point 5. As Figure 8 shown, one red marking point 3, one green marking point 4, and one blue marking point 5 are pasted on the surface of the model, and the three marking points are not collinear. Three RGB one-dimensional cameras are arranged, and a three-dimensional coordinate measuring device is constructed by referring to the method in the prior art "Research on Multi-line Array CCD Space Object Pose Measurement Based on Multi-point Cooperative Target".

[0087] As Figure 9As shown, the RGB three-color one-dimensional camera consists of an RGB three-line array image sensor 6 and a cylindrical lens 12. The cylindrical lens 12 is used to converge the light rays of the marked point in space into an image line 13. The image line 13 is orthogonal to the linear array imaging unit of the RGB three-line array image sensor 6, thereby measuring the one-dimensional coordinates of the marked point in space.

[0088] The RGB three-line array image sensor has an imaging resolution of 4096*3 pixels and a line frequency of 70KHz.

[0089] Red marker 3, green marker 4, and blue marker 5 affixed to the model surface are imaged on an RGB three-line array image sensor, simultaneously acquiring three linear array images (Ir, Ig, and Ib) in the R, G, and B channels. The data bit width of Ir, Ig, and Ib is 8 bits.

[0090] Given t = 200 and s = 3, extract the sub-pixel coordinates Pr, Pg, and Pb of the center of three cooperative marker points from the linear array images of Ir, Ig, and Ib. Referring to the method in the existing technology "Research on Attitude Measurement of Multi-Linear CCD Spatial Objects Based on Multi-Point Cooperative Targets", calculate the three-dimensional coordinates of red marker point 3, green marker point 4, and blue marker point 5. Then, use the method in the existing technology "Research on Attitude Measurement of Multi-Linear CCD Spatial Objects Based on Multi-Point Cooperative Targets" to calculate the model attitude angle.

[0091] Example 3

[0092] The difference from Example 2 is that the red marker 3, green marker 4, and blue marker 5 are made of diodes that emit red, green, and blue light, respectively.

[0093] Example 4

[0094] The difference from Example 2 is that the red marker 3, green marker 4, and blue marker 5 are made of round LED beads that emit red, green, and blue light, respectively.

[0095] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0096] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for online measurement of attitude angles of a high instantaneous velocity model, characterized in that, Said method comprises: arranging n cooperative marker points on the surface of a model under test, wherein n is an integer greater than or equal to 3, any three cooperative marker points are not collinear, each cooperative marker point emits or reflects one of three colors of red, green and blue light, and adjacent two cooperative marker points emit or reflect light with different colors; constructing a three-dimensional coordinate measuring device with three one-dimensional cameras, and measuring the three-dimensional coordinates of the n cooperative marker points synchronously by using the three-dimensional coordinate measuring device; when n=3, solving and obtaining the attitude angle of the model under test according to the three-dimensional coordinates of the n cooperative marker points; when n>3, calculating a transformation matrix of time t relative to time t0 by using a least square method, and calculating and obtaining the attitude angle of the model under test based on the transformation matrix and the three-dimensional coordinates of the n cooperative marker points.

2. The method for online measurement of attitude angle of a high instantaneous velocity model according to claim 1, characterized in that, when n>3, any two cooperative marker points among the n cooperative marker points emit or reflect light with different colors.

3. The method for online measurement of attitude angle of a high instantaneous velocity model according to claim 1, characterized in that, Said one-dimensional camera comprises a cylindrical mirror and an RGB three-line array image sensor, wherein the cylindrical mirror is configured to focus the light of the cooperative marker points on the surface of the model under test onto the RGB three-line array image sensor; the RGB three-line array image sensor comprises three linear array imaging units respectively configured to sense red, green and blue light, and the three linear array imaging units are flush end to end and arranged side by side in parallel.

4. The method for online measurement of attitude angle of a high instantaneous velocity model according to claim 1, characterized in that, The light reflected or emitted by said cooperative marker points presents as circular light spots.

5. The method for online measurement of attitude angle of a high instantaneous velocity model according to claim 1, characterized in that, Said cooperative marker points are made of light-emitting diodes or LED beads, or said cooperative marker points are made into disc patterns with one of red, green and blue paints and adhered to the surface of the model under test, and are illuminated by white light.

6. The method for online measurement of attitude angle of a high instantaneous velocity model according to claim 3, characterized in that, The R, G and B linear array imaging units in said RGB three-line array image sensor respectively image the cooperative marker points to obtain Ir, Ig and Ib linear array images; the central sub-pixel coordinates of the cooperative marker points are extracted and obtained from the Ir, Ig and Ib linear array images, and the three-dimensional coordinates of the cooperative marker points are calculated and obtained based on the central sub-pixel coordinates of the cooperative marker points, the imaging model of the one-dimensional cameras and the geometric positional relationship of the three one-dimensional cameras.

7. The method for online measurement of attitude angle of a high instantaneous velocity model according to claim 6, characterized in that, The specific method for extracting and obtaining the central sub-pixel coordinate Pr of the cooperative marker point from the Ir linear array image is as follows: Step a-1: setting a maximum brightness pixel marking vector Vr, the length of Vr is the same as that of Ir, and the initial values of all elements in Vr are set to 0; Step a-2: according to the marking vector Vr, finding the pixel Rm with the maximum brightness among all pixels marked as 0 in Vr in the Ir linear array image, and setting the element at the position of Rm in Vr to 1; Step a-3: extracting pixels at the position of Rm from the Ig and Ib linear array images to obtain Ig(Rm) and Ib(Rm); when Ig(Rm)<T and Ib(Rm)<T, taking pixel values in a local neighborhood of 2*s+1 pixels with Rm as the center in the Ir linear array image, calculating and obtaining the central sub-pixel coordinate Pr, wherein T is a preset value and s is a window size; when Ig(Rm)>T and Ib(Rm)>T, turning to step a-2.

8. The method for online measurement of attitude angle of a high instantaneous velocity model according to claim 6, characterized in that, The specific method for extracting and obtaining the central sub-pixel coordinate Pg of the cooperative marker point from the Ig linear array image is as follows: Step b-1: set a maximum-brightness pixel marking vector Vg, the length of Vg is the same as that of Ig, and all elements in Vg are initialized to 0; Step b-2: according to the marking vector Vg, find the pixel Gm with the maximum brightness among all pixels marked 0 in Vg in the Ig linear array image, and set the element at the position of Gm in Vg to 1; Step b-3: extract pixels at the position of Gm from Ir and Ib linear array images to obtain Ir(Gm) and Ib(Gm); when Ir(Gm)<T and Ib(Gm)<T, in the Ig linear array image, take pixel values in a local neighborhood of 2*s+1 pixels centered on Gm, calculate and obtain the central sub-pixel coordinate Pg of the cooperative marker point, where T is a preset value and s is the window size; when Ir(Gm)>T and Ib(Gm)>T, go to Step b-2.

9. The method for online measurement of attitude angle of a high instantaneous velocity model according to claim 6, characterized in that, The specific method for extracting and obtaining the central sub-pixel coordinate Pb of the cooperative marker point from the Ib linear array image is as follows: Step c-1: set a maximum-brightness pixel marking vector Vb, the length of Vb is the same as that of Ib, and all elements in Vb are initialized to 0; Step c-2: according to the marking vector Vb, find the pixel Bm with the maximum brightness among all pixels marked 0 in Vb in the Ib linear array image, and set the element at the position of Bm in Vb to 1; Step c-3: extract pixels at the position of Bm from Ir and Ig linear array images to obtain Ir(Bm) and Ig(Bm); when Ir(Bm)<T and Ig(Bm)<T, in the Ib linear array image, take pixel values in a local neighborhood of 2*s+1 pixels centered on Bm, calculate the central sub-pixel coordinate Pb of the cooperative marker point, where T is a preset value and s is the window size; when Ir(Bm)>T and Ig(Bm)>T, go to Step c-2.

Images