Liquid concentration and underwater terrain acquisition method and system based on Kinect

Abstract

The invention belongs to the technical field of non-intrusive underwater collection, and particularly discloses a liquid concentration and underwater terrain collection method and system based on Kinect, and the method comprises the following steps: obtaining water surface point cloud data and underwater terrain point cloud data based on a Kinect camera; the method comprises the following steps of: fitting a relation curve of a gray value and liquid concentration by preparing mixed liquid with different sediment volume concentrations; importing the water surface point cloud data into point cloud visualization software for processing to obtain an RGB value of the measured liquid, and obtaining the liquid concentration of the measured liquid through a relation curve of a gray value and the liquid concentration; correcting the obtained underwater topography point cloud data to obtain corrected point cloud data of the underwater topography; and importing the corrected point cloud data into point cloud visualization software to carry out underwater terrain three-dimensional reconstruction. According to the invention, one-to-one correspondence color maps and depth maps are obtained based on the Kinect camera, the liquid level liquid concentration can be effectively collected, and meanwhile, three-dimensional reconstruction of the underwater terrain can be quickly realized.

Description

Technical Field

[0001] The invention belongs to the technical field of non-invasive underwater data collection, and in particular relates to a method for collecting liquid concentration and underwater terrain based on Kinect. Background Art

[0002] Liquid concentration is an important physical quantity in daily life. It is one of the main parameters to characterize liquid properties and an important measurement parameter in industrial and agricultural production and scientific research experiments. Many physical and chemical properties of substances are easily affected by concentration. Too high or too low concentration will affect the quality of the product. However, in the prior art, fiber grating, ultrasonic and other technologies are mainly used to obtain liquid concentration, but the equipment relying on these technologies is generally expensive and complicated to operate.

[0003] At the same time, the existing underwater surveying and mapping technology is greatly affected by the water environment in optical imaging, and the application of optical imaging technology cannot accurately image in complex water environments; the ordinary underwater engineering surveying and mapping methods on the market are currently two-dimensional surveying and mapping, which are cumbersome to operate and the accuracy of surveying and mapping data is low, and it is impossible to timely detect engineering safety hazards. The technology used for underwater three-dimensional surveying and mapping is currently only used in extremely important projects because of its high cost. Summary of the invention

[0004] The technical problem to be solved by the present invention is to provide a Kinect-based liquid concentration and underwater terrain acquisition method in view of the above-mentioned deficiencies in the prior art, which can simply and efficiently acquire liquid concentration and simultaneously perform three-dimensional reconstruction of underwater terrain.

[0005] The technical solution adopted by the present invention is: a method for collecting liquid concentration and underwater terrain based on Kinect, comprising the following steps:

[0006] Step 1: Acquire a water surface point cloud image and an underwater terrain point cloud image based on a Kinect camera, and acquire water surface point cloud data and underwater terrain point cloud data according to the water surface point cloud image and the underwater terrain point cloud image, wherein the water surface point cloud data is the coordinate value and RGB value of each pixel point in the water surface point cloud image, and the underwater terrain point cloud data is the coordinate value of each pixel point in the underwater terrain point cloud image;

[0007] Step 2: Prepare mixed liquids with different sediment volume concentrations, obtain water surface point cloud data of liquids with different sediment volume concentrations based on the Kinect camera, obtain the grayscale values ​​of liquids with different sediment volume concentrations through the RGB values ​​of the liquids with different sediment volume concentrations, and fit the relationship curve between the grayscale value and the liquid concentration according to the grayscale values ​​of the liquids with different sediment volume concentrations and the sediment volume concentration of the corresponding liquids;

[0008] Step 3: Import the water surface point cloud data into the point cloud visualization software for point cloud segmentation processing, segment the water surface point cloud image with the liquid surface geometric center as the rotation center to obtain the segmented area, perform mode statistics on the RGB value of each pixel in the segmented area, obtain the RGB value of the segmented area as the RGB value of the collected liquid, calculate the gray value according to the RGB value of the collected liquid, and obtain the liquid concentration of the collected liquid according to the gray value and liquid concentration relationship curve in step 2;

[0009] Step 4: Using the differential form of the law of refraction for the acquired underwater terrain point cloud data, derive the function that relates the refraction angle, the incident angle, and the coordinates of the real image point and the virtual image point, perform refraction correction on the underwater terrain point cloud data, and obtain the refraction-corrected point cloud data of the underwater terrain. Then, perform water depth correction and water concentration correction in turn based on the refraction-corrected point cloud data, and finally obtain the corrected point cloud data;

[0010] Step 5: Import the corrected point cloud data into the point cloud visualization software to perform point cloud segmentation, statistical filtering to remove outliers, point cloud secondary sampling and underwater terrain 3D reconstruction.

[0011] Preferably, the Kinect camera in step 1 includes a color camera and a depth sensor, and its use range is controlled between 0.5m-4.5m, wherein the color camera is used to collect RGB values ​​of point clouds in the surface point cloud image and coordinate values ​​of point clouds in the underwater terrain point cloud image, and the depth sensor is used to obtain depth information of the underwater terrain.

[0012] Preferably, in step 2, the gray value is calculated as follows:

[0013]

[0014] Wherein, g represents the gray value of the liquid, R represents the R representative value of the liquid, G represents the G representative value of the liquid, and B represents the B representative value of the liquid;

[0015] The relationship curve equation between gray value and liquid concentration is:

[0016]

[0017] in, Represents the gray value of the liquid, Indicates the concentration of the liquid.

[0018] As a preferred method, in step 4, the underwater terrain point cloud data is corrected, specifically: establish a yoz plane coordinate system, set the water surface as the z-axis, the real image position coordinates of the underwater point S are (y0, z0), the refractive index of water is n, the refractive index of air is n0=1, the elementary rays emitted by point S are SM1 and SM2, the rays refracted by the water surface are M1 P1 and M2 P2, and the incident angle of the ray SM1 is recorded as , the refraction angle is recorded as , the intersection of the reverse extension lines of the light rays M1 P1 and M2 P2 is S1, point S1 is the virtual image position of point S, recorded as (y, z), the point on the y-axis corresponding to point S is recorded as point A, the point on the y-axis corresponding to point S1 is recorded as point B, and the angle between the elementary rays SM1 and SM2 is recorded as , the angle between the reverse extension lines of light M1 P1 and M2 P2 is recorded as ;

[0019] According to the law of refraction,

[0020] ,Right now

[0021] Differentiating both sides gives

[0022] Formula (1):

[0023] In △SM1M2, we can get formula (2):

[0024] In △S1M1M2, we can get formula (3):

[0025] From equation (2) and equation (3), we can get equation (4):

[0026] From equation (1) and equation (4), we can get equation (5):

[0027] In △SAM1, we can get ,Right now

[0028] In △S1BM1, we can get ,Right now

[0029] Will and Substituting into equation (5) we can get equation (6):

[0030] again

[0031] Then we get formula (7):

[0032]

[0033] Formula (8):

[0034] Substituting equation (6) and equation (7) into equation (8), we can get equation (9):

[0035]

[0036] Formula (10): , Substituting equation (7) and equation (9) into equation (10), we can get

[0037]

[0038] That is, after refraction correction of underwater terrain point cloud data, refraction-corrected point cloud data can be obtained.

[0039] Preferably, the water depth correction is to multiply the horizontal coordinate value and the vertical coordinate value in the refraction correction point cloud data by the reduction coefficient R1 respectively to obtain the water depth correction point cloud data;

[0040] The reduction coefficient R1=1.005-0.013h,

[0041] Wherein, R1 is the reduction factor related to water depth; h is the water depth.

[0042] Preferably, the water concentration correction is to multiply the horizontal coordinate value and the vertical coordinate value in the water depth correction point cloud data by the horizontal coordinate error compensation coefficient R2 and the vertical coordinate error compensation coefficient R3 respectively, and finally obtain the correction point cloud data;

[0043] The horizontal coordinate error compensation coefficient R2=0.966-2839.31× ,

[0044] The longitudinal coordinate error compensation coefficient R3=0.937+7744.39× ,

[0045] in, is the liquid concentration.

[0046] A Kinect-based liquid concentration and underwater topography acquisition system, comprising a main control module and a processing module, a data acquisition module, a water concentration processing module, a refraction coordinate correction module, a water depth coordinate correction module and a water concentration coordinate correction module respectively connected to the main control module;

[0047] The data acquisition module is used to collect underwater surface point cloud images and underwater terrain point cloud images;

[0048] The water concentration processing module is used to calculate the liquid concentration of the collection site according to the RGB values ​​in the water surface point cloud data obtained by the processing module;

[0049] The refraction coordinate correction module is used to perform refraction correction according to the coordinate values ​​in the underwater terrain point cloud data obtained by the processing module;

[0050] The water depth coordinate correction module is used to perform water depth correction on the coordinate values ​​corrected by the refraction coordinate correction module, specifically: multiplying the horizontal coordinate value and the vertical coordinate value corrected by the refraction coordinate correction module by a reduction coefficient R1 respectively;

[0051] The reduction coefficient R1=1.005-0.013h,

[0052] Where R1 is the reduction factor related to water depth; h is the water depth;

[0053] The water concentration coordinate correction module is used to compensate and correct the coordinate values ​​corrected by the water depth coordinate correction module, specifically, multiplying the horizontal coordinate value corrected by the water depth coordinate correction module by the horizontal coordinate error compensation coefficient R2, and multiplying the longitudinal coordinate value corrected by the water depth coordinate correction module by the longitudinal coordinate error compensation coefficient R3, and finally obtaining the corrected point cloud data;

[0054] The horizontal coordinate error compensation coefficient R2=0.966-2839.31× ,

[0055] The longitudinal coordinate error compensation coefficient R3=0.937+7744.39× ,

[0056] in, is the liquid concentration;

[0057] The processing module is used to convert the collected underwater surface point cloud images and underwater terrain point cloud images into surface point cloud data and underwater terrain point cloud data, wherein the surface point cloud data is the coordinate value and RGB value of each pixel point in the surface point cloud image, and the underwater terrain point cloud data is the coordinate value of each pixel point in the underwater terrain point cloud image; the corrected point cloud data finally obtained is processed, and the underwater terrain is reconstructed in three dimensions after the processing.

[0058] The beneficial effects of the present invention are:

[0059] (1) The water surface point cloud image is obtained based on the Kinect camera, and the liquid concentration can be collected by combining the relationship curve between liquid concentration and grayscale value;

[0060] (2) Based on the underwater terrain point cloud data obtained by the Kinect camera, the real data of the underwater terrain can be obtained after refraction correction, water depth correction and water concentration correction, so as to realize the three-dimensional reconstruction of the underwater terrain through the point cloud visualization software;

[0061] The present invention obtains one-to-one corresponding color images and depth images based on the Kinect camera, can effectively collect the liquid concentration of the liquid surface, and can quickly realize the three-dimensional reconstruction of the underwater terrain. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 It is a flowchart of the present invention;

[0063] Figure 2 is a relationship curve between liquid concentration and gray value of the present invention;

[0064] Figure 3 This is a schematic diagram of the refraction principle of underwater terrain point cloud data correction in the present invention;

[0065] Figure 4 is the relationship curve between water depth and reduction coefficient of the present invention;

[0066] Figure 5 It is the relationship curve between the liquid concentration and the horizontal coordinate error compensation coefficient and the liquid concentration and the longitudinal coordinate error compensation coefficient of the present invention. DETAILED DESCRIPTION

[0067] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0068] Example

[0069] like Figure 1 As shown, the method for collecting liquid concentration and underwater terrain based on Kinect provided in this embodiment includes the following steps:

[0070] Step 1: A water surface point cloud image and an underwater terrain point cloud image are obtained based on a Kinect camera. The Kinect camera includes a color camera and a depth sensor. A color image is obtained by the color camera, a depth map is obtained by the depth sensor, and a one-to-one correspondence between the color image and the depth map is completed by a program provided by the Kinect camera. That is, the color camera can obtain an image with a resolution of 1920×1080, and the depth sensor can obtain a depth map with a resolution of 512×424. In this embodiment, the Kinect camera adopts Kinect2.0, and its optimal detection range is between 0.5m and 4.5m.

[0071] A water surface point cloud image and an underwater terrain point cloud image are obtained according to the captured color image and depth image, and then surface point cloud data and underwater terrain point cloud data are obtained according to the water surface point cloud image and the underwater terrain point cloud image, wherein the water surface point cloud data is the coordinate value and RGB value of each pixel point in the water surface point cloud image, and the underwater terrain point cloud data is the coordinate value of each pixel point in the underwater terrain point cloud image;

[0072] Step 2: Prepare water-soil mixed solutions with sediment volume concentrations of 0%, 5%, 10%, 15%, 20%, 25%, 30%, and 35% respectively, stir the prepared mixed solutions evenly, and obtain the water surface point cloud data of the mixed solutions with different sediment volume concentrations based on the Kinect camera. The water surface point cloud data contains the X, Y, Z, and RGB information of the point cloud. The water surface point cloud data of the mixed solutions with different sediment volume concentrations are imported into the CloudCompare point cloud visualization software for segmentation processing, that is, the geometric center of the liquid surface is taken as the rotation center, and a cubic point cloud segmentation area of ​​X=0.01m, Y=0.01m, and Z=0.005m is taken. The RGB value of each point cloud in the point cloud segmentation area is counted respectively, and the mode of the R value is taken as the R representative value of the collected mixed liquid, the mode of the G value is taken as the G representative value of the collected mixed liquid, and the mode of the B value is taken as the B representative value of the collected mixed liquid. The obtained R representative value, G representative value, and B representative value are recombined into a new RGB value, and the new RGB value is the RGB value of the corresponding collected mixed liquid;

[0073] According to the gray value calculation method:

[0074] Wherein, g represents the gray value of the liquid, R represents the R representative value of the liquid, G represents the G representative value of the liquid, and B represents the B representative value of the liquid;

[0075] The gray value of the corresponding collected mixed liquid is calculated, and the relationship curve between the gray value and the liquid concentration is obtained by fitting according to the known liquid concentration of the mixed liquid with different sediment volume concentrations and the obtained gray value of the mixed liquid with different sediment volume concentrations, such as Figure 2 As shown, the curve equation of gray value and liquid concentration is:

[0076]

[0077] in, Represents the gray value of the liquid, Indicates liquid concentration;

[0078] Step 3: Import the water surface point cloud data obtained in step 1 into CloudCompare point cloud visualization software for segmentation processing, that is, take the geometric center of the liquid surface as the rotation center, take the cubic point cloud segmentation area of ​​X=0.01m, Y=0.01m, and Z=0.005m, and count the RGB values ​​of each point cloud in the point cloud segmentation area respectively, and take the mode of the R value as the R representative value of the collected liquid, the mode of the G value as the G representative value of the collected liquid, and the mode of the B value as the B representative value of the collected liquid. The obtained R representative value, G representative value, and B representative value are recombined into a new RGB value, and the new RGB value is the RGB value of the corresponding collected liquid. According to the RGB value of the collected liquid and the grayscale calculation method in step 2, the grayscale value of the collected liquid is obtained, and then the liquid concentration of the collected liquid is calculated according to the curve equation of the grayscale value and liquid concentration obtained in step 2;

[0079] Step 4: Figure 3 As shown, the underwater terrain point cloud data is corrected by using the differential form of the refraction law to obtain the function of the refraction angle, the incident angle and the coordinates of the real image point and the virtual image point. Specifically,

[0080] Establish the yoz plane coordinate system, set the water surface as the z-axis, the real image position coordinates of the underwater point S as (y0, z0), the refractive index of water is n, the refractive index of air is n0=1, the elementary rays emitted from point S are SM1 and SM2, the rays refracted by the water surface are M1 P1 and M2 P2, and the incident angle of SM1 is recorded as , the refraction angle is recorded as , the intersection of the reverse extension lines of the light rays M1 P1 and M2 P2 is S1, point S1 is the virtual image position of point S, recorded as (y, z), the point on the y-axis corresponding to point S is recorded as point A, the point on the y-axis corresponding to point S1 is recorded as point B, and the angle between the elementary rays SM1 and SM2 is recorded as , the angle between the reverse extension lines of light M1 P1 and M2 P2 is recorded as ;

[0081] According to the law of refraction,

[0082] ,Right now

[0083] Differentiating both sides gives

[0084] Formula (1):

[0085] In △SM1M2, we can get formula (2):

[0086] In △S1M1M2, we can get formula (3):

[0087] From equation (2) and equation (3), we can get equation (4):

[0088] From equation (1) and equation (4), we can get equation (5):

[0089] In △SAM1, we can get ,Right now

[0090] In △S1BM1, we can get ,Right now

[0091] Will and Substituting into equation (5) we can get equation (6):

[0092] again

[0093] Then we get formula (7):

[0094]

[0095] Formula (8):

[0096] Substituting equation (6) and equation (7) into equation (8), we can get equation (9):

[0097]

[0098] Formula (10): , Substituting equation (7) and equation (9) into equation (10), we can get

[0099]

[0100] That is, the virtual image position S1 (x, y, z) is the position coordinate of the underwater terrain point cloud data acquisition, and the real image position S (x, y0, z0) is the refraction correction position coordinate of the underwater terrain after the virtual image position S1 (x, y, z) is corrected;

[0101] When there are suspended objects in the water, turbid water will cause significant loss of infrared energy for two reasons: first, the water depth has a strong blocking effect on infrared light, which consumes light energy and affects the measurement accuracy of Kinect 2.0; second, suspended objects in the water will block light and reduce the measurement accuracy of Kinect 2.0;

[0102] Since water depth has a strong blocking effect on infrared light, which consumes light energy and affects the measurement accuracy of Kinect 2.0, tests were conducted in clear static water at a certain depth. After refraction correction was performed on the measured horizontal and vertical data, there was still some deviation between the measured values ​​and the actual values;

[0103] like Figure 4 As shown in the figure, different clear and static water depths are selected for fixed-point measurement, and the same point is measured in the absence of water. According to the true coordinate values ​​obtained in the absence of water and the error coordinate values ​​obtained by measuring at different water depths, the relationship curve between water depth and reduction coefficient is fitted according to the empirical formula to obtain the relationship between water depth and reduction coefficient: R1=1.005-0.013h,

[0104] Where R1 is the reduction factor related to water depth; h is the water depth;

[0105] By selecting appropriate water depth detection equipment, the actual water depth of the area to be measured is measured, and the corresponding reduction coefficient is obtained according to the actual water depth. The coordinate values ​​(x1, y1, z1) of the water depth correction point cloud data are obtained by multiplying the refraction-corrected coordinate values ​​(x, y0, z0) by the reduction coefficient R1 obtained according to the actual water depth.

[0106] like Figure 5 As shown in the figure, after the water depth correction, the remaining error can be considered to be caused by muddy water and compensated accordingly. The error coordinate values ​​in different water concentrations are empirically fitted with the true coordinate values ​​to obtain the horizontal coordinate error compensation coefficient R2 and the longitudinal coordinate error compensation coefficient R3 caused by the liquid concentration.

[0107] The horizontal coordinate error compensation coefficient R2=0.966-2839.31× ,

[0108] The longitudinal coordinate error compensation coefficient R3=0.937+7744.39× ,

[0109] in, is the liquid concentration;

[0110] According to the coordinate values ​​(x1, y1, z1) in the depth correction point cloud data, the horizontal coordinate values ​​(x1, y1) are multiplied by the horizontal error compensation coefficient R2, and the longitudinal coordinate value z1 is multiplied by the longitudinal coordinate error compensation coefficient R3, so as to obtain the correction point cloud data of the underwater terrain;

[0111] The longitudinal coordinate value in step 4 is the height coordinate value perpendicular to the water surface;

[0112] Step 5: Import the corrected point cloud data into CloudCompare point cloud visualization software, and use the point cloud cutting tool in CloudCompare point cloud visualization software to cut out all point clouds except the dam body, so that only the corrected point cloud data of the dam body is retained, and then the statistical filtering module in CloudCompare point cloud visualization software is used to remove outliers. The outliers that are not automatically removed by the statistical filtering module can be removed again using the cutting tool, and then the quadratic surface is selected through the surface normal vector calculation model in CloudCompare point cloud visualization software. Finally, the PoissonRecon plug-in in CloudCompare point cloud visualization software is used to fit the overall surface of the dam body, that is, to achieve three-dimensional reconstruction of the dam body in the underwater terrain; the three-dimensionally reconstructed dam body in the underwater terrain is used to facilitate the subsequent analysis of the dam body breach.

[0113] The above description is only a preferred implementation manner of the present invention, but the protection scope of the present invention is not limited thereto, and any modification and replacement based on the technical solution and inventive concept provided by the present invention should be included in the protection scope of the present invention.

Claims

1. A method for collecting liquid concentration and underwater terrain based on Kinect, characterized in that: The following steps are involved: Step 1: Acquire a water surface point cloud image and an underwater terrain point cloud image based on a Kinect camera, and acquire water surface point cloud data and underwater terrain point cloud data according to the water surface point cloud image and the underwater terrain point cloud image, wherein the water surface point cloud data is the coordinate value and RGB value of each pixel point in the water surface point cloud image, and the underwater terrain point cloud data is the coordinate value of each pixel point in the underwater terrain point cloud image; Step 2: Prepare mixed liquids with different sediment volume concentrations, obtain water surface point cloud data of liquids with different sediment volume concentrations based on the Kinect camera, obtain the grayscale values ​​of liquids with different sediment volume concentrations through the RGB values ​​of the liquids with different sediment volume concentrations, and fit the relationship curve between the grayscale value and the liquid concentration according to the grayscale values ​​of the liquids with different sediment volume concentrations and the sediment volume concentration of the corresponding liquids; Step 3: Import the water surface point cloud data into the point cloud visualization software for point cloud segmentation processing, segment the water surface point cloud image with the liquid surface geometric center as the rotation center to obtain the segmented area, perform mode statistics on the RGB value of each pixel in the segmented area, obtain the RGB value of the segmented area as the RGB value of the collected liquid, calculate the gray value according to the RGB value of the collected liquid, and obtain the liquid concentration of the collected liquid according to the gray value and liquid concentration relationship curve in step 2; Step 4: Using the differential form of the law of refraction for the acquired underwater terrain point cloud data, derive the function that relates the refraction angle, the incident angle, and the coordinates of the real image point and the virtual image point, perform refraction correction on the underwater terrain point cloud data, and obtain the refraction-corrected point cloud data of the underwater terrain. Then, perform water depth correction and water concentration correction in turn based on the refraction-corrected point cloud data, and finally obtain the corrected point cloud data; Step 5: Import the corrected point cloud data into the point cloud visualization software to perform point cloud segmentation, statistical filtering to remove outliers, point cloud secondary sampling and underwater terrain 3D reconstruction.

2. The method for collecting liquid concentration and underwater terrain based on Kinect according to claim 1, characterized in that: The Kinect camera in step 1 includes a color camera and a depth sensor, and its use range is controlled between 0.5m and 4.5m. The color camera is used to collect the RGB values ​​of the point cloud in the surface point cloud image and the coordinate values ​​of the point cloud in the underwater terrain point cloud image, and the depth sensor is used to obtain the depth information of the underwater terrain.

3. The method for collecting liquid concentration and underwater terrain based on Kinect according to claim 2, characterized in that: In step 2, the gray value is calculated as follows: ， Wherein, g represents the gray value of the liquid, R represents the R representative value of the liquid, G represents the G representative value of the liquid, and B represents the B representative value of the liquid; The relationship curve equation between gray value and liquid concentration is: ， in, Represents the gray value of the liquid, Indicates the concentration of the liquid.

4. The method for collecting liquid concentration and underwater terrain based on Kinect according to claim 3, characterized in that: In step 4, the underwater terrain point cloud data is refraction corrected. Specifically, the yoz plane coordinate system is established, the water surface is set as the z-axis, the real image position coordinates of the underwater point S are (y0, z0), the refractive index of water is n, the refractive index of air is n0=1, the elementary rays emitted by point S are SM1 and SM2, the rays refracted by the water surface are M1 P1 and M2 P2, and the incident angle of SM1 is recorded as , the refraction angle is recorded as , the intersection of the reverse extension lines of light M1 P1 and M2 P2 is S1, point S1 is the position of the virtual image of point S, recorded as (y, z), According to the law of refraction, ,Right now , Right now: ， ， That is, after refraction correction of underwater terrain point cloud data, refraction-corrected point cloud data can be obtained.

5. The method for collecting liquid concentration and underwater terrain based on Kinect according to claim 4, characterized in that: The water depth correction is to multiply the horizontal coordinate value and the vertical coordinate value in the refraction correction point cloud data by the reduction coefficient R1, that is, to obtain the water depth correction point cloud data; The reduction coefficient R1=1.005-0.013h, Wherein, R1 is the reduction factor related to water depth; h is the water depth.

6. The method for collecting liquid concentration and underwater terrain based on Kinect according to claim 5, characterized in that: The water concentration correction is to multiply the horizontal coordinate value and the vertical coordinate value in the water depth correction point cloud data by the horizontal coordinate error compensation coefficient R2 and the vertical coordinate error compensation coefficient R3 respectively, and finally obtain the correction point cloud data; The horizontal coordinate error compensation coefficient R2=0.966-2839.31× , The longitudinal coordinate error compensation coefficient R3=0.937+7744.39× , in, is the liquid concentration.

7. A collection system for a liquid concentration and underwater terrain collection method based on Kinect according to any one of claims 1 to 6, characterized in that: It includes a main control module and a processing module, a data acquisition module, a water concentration processing module, a refraction coordinate correction module, a water depth coordinate correction module and a water concentration coordinate correction module respectively connected to the main control module; The data acquisition module is used to collect underwater surface point cloud images and underwater terrain point cloud images; The water concentration processing module is used to calculate the liquid concentration of the collection site according to the RGB value in the water surface point cloud data obtained by the processing module; The refraction coordinate correction module is used to perform refraction correction according to the coordinate values ​​in the underwater terrain point cloud data obtained by the processing module; The water depth coordinate correction module is used to perform water depth correction on the coordinate values ​​corrected by the refraction coordinate correction module, specifically: multiplying the horizontal coordinate value and the vertical coordinate value corrected by the refraction coordinate correction module by a reduction coefficient R1 respectively; The reduction coefficient R1=1.005-0.013h, Where R1 is the reduction factor related to water depth; h is the water depth; The water concentration coordinate correction module is used to compensate and correct the coordinate values ​​corrected by the water depth coordinate correction module, specifically, multiplying the horizontal coordinate value corrected by the water depth coordinate correction module by the horizontal coordinate error compensation coefficient R2, and multiplying the longitudinal coordinate value corrected by the water depth coordinate correction module by the longitudinal coordinate error compensation coefficient R3, and finally obtaining the corrected point cloud data; The horizontal coordinate error compensation coefficient R2=0.966-2839.31× , The longitudinal coordinate error compensation coefficient R3=0.937+7744.39× , in, is the liquid concentration; The processing module is used to convert the collected underwater surface point cloud images and underwater terrain point cloud images into surface point cloud data and underwater terrain point cloud data, wherein the surface point cloud data is the coordinate value and RGB value of each pixel point in the surface point cloud image, and the underwater terrain point cloud data is the coordinate value of each pixel point in the underwater terrain point cloud image; the corrected point cloud data finally obtained is processed, and the underwater terrain is reconstructed in three dimensions after the processing.

Images