CN110645921A - A three-dimensional measurement method of ice shape based on polarization imaging - Google Patents

Abstract

The invention proposes a three-dimensional measurement method of ice shapes based on polarization imaging. A linear polarized light source is used to illuminate various ice shapes, a polarization camera shoots a laser light bar, and a rotating platform scans the icing surface; Calibration is performed; the rotary table is calibrated; the center line of the laser light bar is extracted by the image processing method, and the three-dimensional coordinates of the center line of the laser light bar in the world coordinate system are calculated; the method of the invention lays a good foundation for the online three-dimensional measurement of ice shape. Theoretical and technical foundations.

Description

A three-dimensional measurement method of ice shape based on polarization imaging

technical field

The invention relates to the fields of optical three-dimensional measurement and computer vision, in particular to an ice-shaped three-dimensional measurement method based on polarization imaging.

Background technique

Aircraft icing widely exists in flight practice and poses a serious threat to flight safety. It will not only destroy the smooth flow of air and increase flight resistance, but also change the aerodynamic characteristics of the wing, causing failure or counter-effect of rudder control. Icing on different parts of the aircraft can affect the safe flight of the aircraft. Almost all icing research now begins by measuring the shape of icing on the surface of an aircraft in an icing wind tunnel.

So far, the methods for measuring the shape of aircraft icing mainly include: cross-sectional profile method, photogrammetry method, casting method, and laser line scanning method. All of the above methods have their advantages, but they also have limitations. In China, Gong Xiaoliang proposed a visible laser three-dimensional scanning method, and for the first time measured the continuous growth of the three-dimensional shape of frost and ice and the three-dimensional shape of bright ice after frost. He also used ice to exhibit Lambertian radiation characteristics in a specific infrared band. An infrared laser three-dimensional scanning method was developed, and the continuous growth of the three-dimensional shape of the ice was measured for the first time. However, this measurement method is essentially a heating method, and it is unknown whether it has an effect on the growth process of the ice shape.

From the technical development trend at home and abroad, optical non-contact online ice shape measurement is the future development trend. However, there is currently no suitable method to achieve online ice shape measurement. Scholars at home and abroad have carried out a lot of research on the improvement of the calibration accuracy of the laser knife cutting method. The projection error of the calibration point in the camera calibration reaches one hundredth of a pixel level, and the laser plane equation can be accurately obtained by the optimization method. However, the laser centerline extraction error still restricts the measurement accuracy. This is due to the high transparency and smooth surface of the ice body, and the strong laser penetration ability. The line laser is projected on the surface of the ice body. Most rays are projected on the ice body, and only a small part of the light rays are projected. Reflection through the surface of the ice body causes serious diffusion of the laser band area of the captured image, which appears as a bright spot area on the image. Therefore, how to filter out the influence of scattered stray light is an urgent problem that needs to be solved in the process of realizing ice shape online measurement.

SUMMARY OF THE INVENTION

In order to solve the above technical problems, the present invention proposes a three-dimensional ice shape measurement method based on polarization imaging, which uses a linearly polarized polarized light source to illuminate various ice shapes so as to perform three-dimensional ice shape measurement. Specifically include the following steps:

Step 1: Use a linearly polarized polarized light source to illuminate the ice, a polarized camera shoots a laser light bar, and the rotating gantry scans the icing surface;

Step 2: Line laser vision sensor calibration;

Step 3: Calibration of the rotary table;

Step 4: Extract the center line of the laser light bar by image processing method, and calculate the three-dimensional coordinates of the center line of the laser light bar in the world coordinate system.

Among them, in step 2, the camera and the line laser together form a line laser vision sensor; its working process is as follows: the line laser projects a light plane to the object to be measured, and the light plane intersects the surface of the object to be measured to generate a light bar, and the light bar contains the object. The height information of the object is then recorded by the camera and transmitted to the computer for analysis to calculate the height information of the object.

Among them, set O _W -X _W Y _W Z _W as the world coordinate system and at the same time as the measurement coordinate system; O _C -X _C Y _C Z _C as the camera coordinate system; O _u -x _u y _u represents the image coordinate system, The unit is pixel; π represents the light plane of structured light; any point P on the light plane is taken, and its coordinates in the measurement coordinate system are P _w =[x _w , y _w , z _w ] ^T , and the projected point on the image plane The image homogeneous coordinates of p are p=[u,v,1] ^T , and the camera imaging model is expressed as:

The measurement coordinate system is established on the camera coordinate system, namely P _w =P _c =[x _c , y _c , z _c ] ^T , then the camera imaging model is expressed as:

Among them: s is the scale coefficient; A is the camera internal parameter matrix, f _u , f _v are the focal length parameters, u ₀ , v ₀ are the main point image coordinates; at the same time, because P is a point on the light plane, the following light plane needs to be satisfied equation:

ax _w +by _w +cz _w +d=0 (3)

Where: a, b, c, d are the coefficients of the light plane equation, respectively.

Among them, by calibrating the camera internal parameter matrix A and the light plane equation in the camera coordinate system, and combining formulas (2) and (3), the three-dimensional coordinates of the point P on the center of the light bar in the camera coordinate system are solved, as shown below:

Among them, considering the lens distortion of the camera, the camera internal parameter matrix A is calibrated by Zhang Zhengyou's calibration method, and the checkerboard target is used for calibration.

Among them, the calibration of the light plane equation using the checkerboard target includes the following four steps:

(1) Target attitude acquisition; place the designed target in the field of view of the camera, the camera shoots the target image, extracts the target feature points, establishes the mapping equation between the image coordinates and the world coordinates, and solves the external parameters of the target, That is, the rotation matrix R and the translation vector T between the camera coordinate system and the target coordinate system;

(2) Extraction of the center of the light bar; keep the target posture unchanged, project the line laser to the surface of the target, shoot the light bar image, and obtain the image coordinates of the light bar center through the light bar center extraction algorithm;

(3) Solve the three-dimensional coordinates of the light bar; determine the three-dimensional coordinates of the light bar center in the camera coordinate system through the target attitude parameters R and T, and the image coordinates of the light bar center;

(4) Solving the light plane; repeating the above operations, taking images of the target and light bar in different postures, and obtaining the three-dimensional coordinates of the center of the light bar under different postures, these coordinates all satisfy the light plane equation, and the coefficients of the light plane equation are determined by the least square method .

Among them, the algorithm for extracting the center of the light strip adopts the gray-scale centroid method.

Among them, in step 3, the camera and the line laser are fixed on the same bracket, and then the turntable is rotated with the bracket to keep the relative position of the camera and the line laser unchanged, that is, the coefficient of the light plane in the camera coordinate system is fixed. ; But the camera rotates with the bracket, so the camera coordinate system is not fixed, it is necessary to calibrate the attitude change of the camera under different rotation angles, and perform the calibration of the rotary table.

Among them, the calibration process of the rotary table includes:

During the rotation of the bracket, the target position is fixed, and two target images are obtained by shooting at angle 1 and angle 2 respectively. According to the calibration of the camera's internal parameters, the external parameter equation between the polarization camera and the target is obtained:

Then the positional relationship of the polarization camera at angle 1 and angle 2 is:

At this time, the camera coordinate systems at different rotation angles are set to the same global coordinate system.

Different from the prior art, the polarized imaging-based three-dimensional ice shape measurement method of the present invention uses a linearly polarized light source to illuminate various ice shapes, a polarized camera shoots a laser light bar, and a rotating gantry scans the icing surface; The laser vision sensor is calibrated; the rotary table is calibrated; the center line of the laser light bar is extracted by the image processing method, and the three-dimensional coordinates of the center line of the laser light bar in the world coordinate system are calculated; by the method of the present invention, the three-dimensional online measurement of ice shape is carried out. Lay a good theoretical and technical foundation.

Description of drawings

Fig. 1: is the method flow chart of the present invention

Figure 2: It is a block diagram of the measurement system structure;

Figure 3: Schematic diagram of line laser vision sensor calibration;

Figure 4: Schematic diagram of the principle of rotation calibration.

Detailed ways

In order to facilitate the understanding and implementation of the present invention by those of ordinary skill in the art, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the embodiments described herein are only used to illustrate and explain the present invention, but not to limit it. this invention.

In the experiment of the present invention, three-dimensional ice shape online measurement is performed in a 3-meter×2-meter icing wind tunnel. The main equipment includes: computer, polarization camera, rotating gantry, line laser, 3m x 2m icing wind tunnel, and ice cubes. The ice cubes are placed in the icing wind tunnel. In order to reduce the interference of the wind tunnel experiment and avoid the influence of low temperature, water vapor and sublimation on the measurement equipment in the icing wind tunnel, cameras, lasers and corresponding optical instruments are installed in the icing wind tunnel. Outside the test section of the wind tunnel, measurements were made through a light-transmitting glass window. The laser and polarization camera are fixed on the rotating gantry, keeping their relative positions unchanged. The icing surface is scanned by the rotating gantry, and the 3D topography data of the entire icing ice shape is obtained by measuring. In the system measurement process, various devices need to be synchronized, such as the synchronization of the camera and the rotary table. The invention adopts the FPGA to read the rotation angle of the rotary table (code disc reading), and uses it as a trigger signal to transmit to the camera for position trigger measurement, and the computer stores the image information of the corresponding position.

Embodiments of the present invention will be described below with reference to FIGS. 1 to 3 . FIG. 1 is a schematic flowchart of a method involved in the present invention, and FIG. 2 is a schematic structural diagram of a device corresponding to implementing the method. The specific steps of the embodiment of the present invention are:

Step 1: Use a linearly polarized polarized light source to illuminate the ice, a polarized camera shoots a laser light bar, and the rotating gantry scans the icing surface;

Step 2: Line laser vision sensor calibration;

In step 2, the camera and the line laser together constitute a line laser vision sensor. As shown in Figure 3, the working process is as follows: the line laser projects a light plane to the measured object, the light plane intersects with the surface of the measured object to generate a light bar, the light bar contains the height information of the object, and then the camera records the light bar image, And transmit it to the computer for analysis, and calculate the height information of the object. Suppose O _W -X _W Y _W Z _W represents the world coordinate system, which is also the measurement coordinate system; O _C -X _C Y _C Z _C represents the camera coordinate system; O _u -x _u y _u represents the image coordinate system, and its unit is pixel ; π represents the light plane of structured light. Take a point P on the light plane arbitrarily, its coordinates in the measurement coordinate system are P _w =[x _w , y _w , z _w ] ^T , and the image homogeneous coordinates of the projected point p on the image plane are p = [u, v,1] ^T .

Usually, we establish the measurement coordinate system on the camera coordinate system, that is, P _w =P _c =[x _c , y _c , z _c ] ^T , then the camera imaging model is:

Among them: s is the scale coefficient; A is the camera internal parameter matrix, f _u , f _v are the focal length parameters, and u ₀ , v ₀ are the coordinates of the main point image. At the same time, because P is a point on the light plane, the following light plane equation needs to be satisfied:

ax _w +by _w +cz _w +d=0 (3)

Where: a, b, c, d are the coefficients of the light plane equation, respectively. Therefore, we only need to calibrate the camera internal parameter matrix A and the light plane equation in the camera coordinate system, and combine the above two formulas to solve the three-dimensional coordinates of a point P on the center of the light bar in the camera coordinate system, as shown below :

In general, the lens distortion of the camera also needs to be considered. The camera internal parameter matrix A can be calibrated by the famous Zhang Zhengyou calibration method, which is usually calibrated with a checkerboard target. The calibration of the light plane equation is solved through the following four steps:

(1) Target attitude acquisition. The designed target is placed in the camera's field of view, the camera captures the target image, extracts the target feature points, and then establishes the mapping equation between the image coordinates and the world coordinates, and then the external parameters of the target can be solved, that is, the camera coordinate system and the Rotation matrix R and translation vector T between target coordinate systems.

(2) Extraction of the center of the light strip. Keeping the target posture unchanged, project the line laser to the target surface, take a light bar image, and obtain the image coordinates of the light bar center through the light bar center extraction algorithm, such as the gray-scale centroid method.

(3) Solve the three-dimensional coordinates of the light bar. Through the target attitude parameters R and T, and the image coordinates of the center of the light bar, the three-dimensional coordinates of the center of the light bar in the camera coordinate system can be determined.

(4) Light plane solution. Repeat the above operations to take images of the target and light bar in different postures, and obtain the three-dimensional coordinates of the center of the light bar in different postures. These coordinates all satisfy the light plane equation, and the coefficients of the light plane equation can be determined by the least squares method.

Step 3: Calibration of the rotary table.

As shown in Figure 4, in step 3, the camera and the line laser are fixed on the same bracket, and then the turntable rotates with the bracket, so the relative position of the camera and the line laser remains unchanged, that is, the coefficient of the light plane in the camera coordinate system is fixed Invariant, therefore different rotation angles, no need to repeat the calibration of the light plane. However, the camera rotates with the bracket, so the camera coordinate system is not fixed, so we need to calibrate the attitude changes of the camera under different rotation angles, which we call rotation table calibration.

During the rotation process, the target position is fixed. We shoot at angle 1 and angle 2 to obtain two target images. According to the calibration of the camera's internal parameters, we can obtain the external parameter equation between the polarization camera and the target:

Then the positional relationship of the polarized camera at angle 1 and angle 2 is:

In this way, we can convert the camera coordinate system under different rotation angles to the same global coordinate system.

Step 4: Extract the center line of the laser light bar by image processing method, and calculate the three-dimensional coordinates of the center line of the laser light bar in the world coordinate system.

Different from the prior art, the polarized imaging-based three-dimensional ice shape measurement method of the present invention uses a linearly polarized light source to illuminate various ice shapes, a polarized camera shoots a laser light bar, and a rotating gantry scans the icing surface; The laser vision sensor is calibrated; the rotary table is calibrated; the center line of the laser light bar is extracted by the image processing method, and the three-dimensional coordinates of the center line of the laser light bar in the world coordinate system are calculated; by the method of the present invention, the three-dimensional online measurement of ice shape is carried out. Lay a good theoretical and technical foundation.

It should be understood that although this specification is described in terms of embodiments, not every embodiment only includes an independent technical solution, and this description in the specification is only for the sake of clarity, and those skilled in the art should take the specification as a whole, and each The technical solutions in the embodiments can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

The series of detailed descriptions listed above are only specific descriptions for the feasible embodiments of the present invention, and they are not used to limit the protection scope of the present invention. Any equivalent embodiments made without departing from the technical spirit of the present invention or Changes should be included within the protection scope of the present invention.

Claims (9)

1. an ice-shaped three-dimensional measurement method based on polarization imaging, is characterized in that, comprises the following steps: Step 1: Use a linearly polarized polarized light source to illuminate the ice, a polarized camera shoots a laser light bar, and the rotating gantry scans the icing surface; Step 2: Line laser vision sensor calibration; Step 3: Calibration of the rotary table; Step 4: Extract the center line of the laser light bar by image processing method, and calculate the three-dimensional coordinates of the center line of the laser light bar in the world coordinate system. 2. a kind of ice-shaped three-dimensional measurement method based on polarization imaging according to claim 1, is characterized in that: in step 2, camera and line laser form together line laser vision sensor; Its working process is as follows: line laser projects light plane To the object to be measured, the light plane intersects with the surface of the object to be measured to generate a light bar, the light bar contains the height information of the object, and then the camera records the light bar image and transmits it to the computer for analysis to calculate the height information of the object. 3. a kind of ice-shaped three-dimensional measuring method based on polarization imaging according to claim 2, is characterized in that: set O _W -X _W Y _W Z _W as world coordinate system, as measuring coordinate system simultaneously; O _C - X _C Y _C Z _C is used as the camera coordinate system; O _u -x _u y _u represents the image coordinate system, and its unit is pixel; π represents the structured light plane; arbitrarily take a point P on the light plane, which is in the measurement coordinate system. The coordinates are P _w =[x _w , y _w , z _w ] ^T , the image homogeneous coordinates of the projected point p on the image plane are p = [u, v, 1] ^T , and the camera imaging model is expressed as:

The measurement coordinate system is established on the camera coordinate system, namely P _w =P _c =[x _c , y _c , z _c ] ^T , then the camera imaging model is expressed as:

Among them: s is the scale coefficient; A is the camera internal parameter matrix, f _u , f _v are the focal length parameters, u ₀ , v ₀ are the main point image coordinates; at the same time, because P is a point on the light plane, the following light plane needs to be satisfied equation: ax _w +by _w +cz _w +d=0 (3) Where: a, b, c, d are the coefficients of the light plane equation, respectively. 4. a kind of ice-shaped three-dimensional measurement method based on polarization imaging according to claim 3 is characterized in that: by calibrating the light plane equation under the camera internal parameter matrix A and the camera coordinate system, the simultaneous formulas (2) and ( 3), solve the three-dimensional coordinates of the point P on the center of the light bar in the camera coordinate system, as shown below:

5. A kind of ice shape three-dimensional measurement method based on polarization imaging according to claim 4, it is characterized in that: considering the lens distortion of the camera, the camera internal parameter matrix A is calibrated by Zhang Zhengyou calibration method, and the checkerboard target is used to calibrate . 6. A kind of ice-shaped three-dimensional measurement method based on polarization imaging according to claim 5, is characterized in that: adopting checkerboard target to demarcate light plane equation comprises following four steps: (1) Target attitude acquisition; place the designed target in the field of view of the camera, the camera shoots the target image, extracts the target feature points, establishes the mapping equation between the image coordinates and the world coordinates, and solves the external parameters of the target, That is, the rotation matrix R and the translation vector T between the camera coordinate system and the target coordinate system; (2) Extraction of the center of the light bar; keep the target posture unchanged, project the line laser to the surface of the target, shoot the light bar image, and obtain the image coordinates of the light bar center through the light bar center extraction algorithm; (3) Solve the three-dimensional coordinates of the light bar; determine the three-dimensional coordinates of the light bar center in the camera coordinate system through the target attitude parameters R and T, and the image coordinates of the light bar center; (4) Solving the light plane; repeating the above operations, taking images of the target and light bar in different postures, and obtaining the three-dimensional coordinates of the center of the light bar under different postures, these coordinates all satisfy the light plane equation, and the coefficients of the light plane equation are determined by the least square method . 7 . The three-dimensional ice shape measurement method based on polarization imaging according to claim 6 , wherein the extraction algorithm of the center of the light strip adopts the gray-scale centroid method. 8 . 8. The ice-shaped three-dimensional measurement method based on polarization imaging according to claim 1, wherein in step 3, the camera and the line laser are fixed on the same bracket, and then the turntable rotates with the bracket to make the camera and the line laser rotate. The relative position of the line laser remains unchanged, that is, the coefficient of the light plane in the camera coordinate system is fixed; but the camera rotates with the bracket, so the camera coordinate system is not fixed, and it is necessary to calibrate the attitude of the camera under different rotation angles change, and perform rotary table calibration. 9 . The ice-shaped three-dimensional measurement method based on polarization imaging according to claim 8 , wherein the process of calibrating the rotating stage comprises: 10 . During the rotation of the bracket, the target position is fixed, and two target images are obtained by shooting at angle 1 and angle 2 respectively. According to the calibration of the camera's internal parameters, the external parameter equation between the polarization camera and the target is obtained:

Then the positional relationship of the polarization camera at angle 1 and angle 2 is:

At this time, the camera coordinate systems at different rotation angles are set to the same global coordinate system.

Images