Cloud height measuring device and method based on ground-based cloud picture

Abstract

The invention discloses a cloud height measuring device and method based on a foundation cloud picture. The cloud height measuring device comprises a plurality of foundation cloud picture acquisition devices and a cloud height measuring device. The n foundation cloud picture acquisition devices acquire foundation cloud picture images from the ground in real time and acquire focal lengths corresponding to different foundation cloud pictures and attitude angles of the different foundation cloud pictures in real time, and the cloud height measurement device processes the foundation cloud pictures acquired by the multiple foundation cloud picture acquisition devices to obtain height information of a target cloud block. The cloud height is measured, and the cost is low; a plurality of foundation cloud picture devices at multiple positions can be added according to requirements, and the measurement range can be flexibly changed; different cloud segmentation methods and cloud block recognition and matching methods are adopted for the characteristics of different cloud layers and cloud blocks in different geographic areas, multi-area popularization is facilitated, and later manual measurement investment is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of meteorological monitoring, and in particular to a cloud height measurement device and method based on ground-based cloud images. Background Art

[0002] Cloud height information is gaining more and more attention and application in fields such as weather forecasting and new energy power generation.

[0003] There are currently three main methods for measuring cloud height: ceilometers, cloud curtain lights, and aircraft such as weather balloons. Ceilometers measure cloud height with high accuracy, but the instrument is complex and the measurement cost is high, which is not conducive to large-scale promotion and use; cloud curtain lights are limited by their principles and can only be used at night; aircraft such as weather balloons measure cloud height, which has a lower cost but is greatly affected by air pressure and wind speed, and has poorer measurement accuracy.

[0004] Binocular ranging technology is an important branch of computer vision that is widely studied and applied. Its main principle is to measure distance by imitating human eyes. It uses two cameras / full-sky imaging devices with the same parameters but in different positions to replace human eyes, collect images of the same target, and calculate the distance information of the target by the difference in the position of the target in the two imaging planes. The target image obtained by the computer is more accurate and easier to process than the information directly obtained by the human eye. Summary of the invention

[0005] Purpose of the invention: The present invention provides a cloud height measurement device and method based on ground-based cloud images, which uses multiple ground-based cloud images to obtain data and performs multiple processing on the obtained data through computer vision methods, and finally obtains the cloud block height information in the ground-based cloud image.

[0006] Technical solution: The cloud height measurement device based on ground-based cloud images described in the present invention includes: multiple ground-based cloud image acquisition devices and cloud height measurement devices; n ground-based cloud image acquisition devices acquire ground-based cloud image images from the ground in real time, and obtain the focal lengths corresponding to different ground-based cloud images and the attitude angles of different ground-based cloud images in real time; the cloud height measurement device processes the ground-based cloud images acquired by the multiple ground-based cloud image acquisition devices to obtain the height information of the target cloud block.

[0007] Furthermore, the cloud height measurement equipment includes a ground-based cloud map image distortion processing module, which is used to perform distortion processing on two or more cloud map images collected at the current sampling moment; a ground-based cloud map image matching module, which is used to match the same target cloud block in two or more ground-based cloud maps to obtain the positions of the edge pixels and the geometric center pixels of the cloud blocks; a ground-based cloud map coordinate system conversion module, which performs coordinate system conversion in the form of matrix conversion according to the focal lengths of different base cloud maps and the attitude angles of the ground-based cloud map acquisition device. The coordinate system conversion is used to convert the pixel positions of the target cloud block pixels via the pixel coordinate system to the image coordinate system, and then from the image coordinate system to the camera coordinate system, and finally from the camera coordinate system to the world coordinate system; a cloud height calculation module, which calculates the target cloud heights of the multiple cloud maps after the coordinate system conversion by using the principle of epipolar geometry in binocular vision and combining the focal length parameters corresponding to the multiple cloud maps.

[0008] Accordingly, a cloud height measurement method based on ground-based cloud images includes the following steps:

[0009] Step 1: Input the multi-ground cloud image, the corresponding focal length f, the lens distortion coefficient, the lens attitude angle, and the lens position information into the cloud height calculation module as parameters for subsequent ground-based cloud image fisheye distortion processing, cloud block identification and matching, coordinate system conversion, and epipolar geometry cloud height calculation;

[0010] Step 2: The ground-based cloud image distortion processing module performs distortion correction on the ground-based cloud image images of the same size and overlapped with each other collected by multiple base cloud image collection devices;

[0011] Step 3: The ground-based cloud image matching module identifies and matches the target cloud blocks in the two ground-based cloud images:

[0012] Step 4: The ground-based cloud map coordinate system conversion module converts the camera coordinate system and the world coordinate system according to specific needs;

[0013] Step 5: The cloud height calculation module calculates the target cloud height using the epipolar geometry principle.

[0014] Furthermore, in step 2, the ground-based cloud image distortion processing module measures the lens distortion coefficient of the ground-based cloud image in advance by a checkerboard measurement method, and then corrects the image by an existing dot matrix calibration method, and only corrects the area of ​​interest or the required area.

[0015] Furthermore, the lens distortion coefficient is measured by a checkerboard calibration method. After the lens distortion coefficient of each ground-based cloud image acquisition device is obtained by the checkerboard calibration method, a corrected image is obtained by an image distortion correction method.

[0016] Furthermore, in step 3, the ground-based cloud image matching module specifically includes the following steps to identify and match the target cloud blocks in the two ground-based cloud images:

[0017] Step 31, segmenting the ground-based cloud image into multiple parts including cloud, sky and sun by using the red-to-blue ratio method;

[0018] Step 32: Identify the same cloud block in the two ground-based cloud images through ORB key point matching, and record the pixel coordinates (X1, Y1) and (X2, Y2) of the cloud block in the two images.

[0019] Step 33: If the same cloud block cannot be identified in step 32, the same cloud block in the two ground-based cloud images is identified through the relationship between the target cloud block and the sun in the two ground-based cloud images, and the pixel coordinates (X1, Y1) and (X2, Y2) of the cloud block in the two images are recorded.

[0020] Furthermore, in step 4, the world coordinate system is the absolute coordinate system of the real three-dimensional space, and the world coordinate system is converted to the camera coordinate system through the rigid body transformation of rotation and translation; the rigid body transformation only changes the spatial position (translation) and orientation (rotation) of the object, but does not change its shape, and is described by two variables: the orthogonal unit rotation matrix R, the three-dimensional translation vector t, and the origin of the world coordinate system is moved to the origin of the camera coordinate system after translation; the rotation has three degrees of freedom, that is, rotation around the three coordinate axes, and the rotation matrices Rx, Ry, and Rz in the three directions can be obtained according to the rotation angle, and the rotation matrix is ​​their product: R=Rx*Ry*Rz; different images should be converted to the corresponding ground-based cloud map device coordinate system.

[0021] Furthermore, in step 5, the cloud height calculation module calculates the target cloud height by using the principle of epipolar geometry. Points CL and CR represent two ground-based cloud image acquisition devices at different locations. The world coordinate system OXYZ is precisely transformed and coincides with the camera coordinate system of the CL ground-based cloud image acquisition device. Points PL and PR represent the imaging points of the target cloud block P in the two cloud images, respectively. The line segment b is the distance between the two ground-based cloud image devices. The triangle PC L CR is similar to triangle PPLPR, and according to the principle of similar triangles, Then the cloud height is derived

[0022] Beneficial effects: Compared with the prior art, the present invention has the following significant advantages: the present invention measures the cloud height, and multiple ground-based cloud image devices can be added at multiple locations according to needs, the measurement range can be flexibly changed, and no more external devices are required, which has the advantages of low cost and good promotion; and compared with methods such as ceilometers and weather balloons, ground-based cloud images have the advantage of sky image visualization, and different cloud segmentation methods and cloud block identification and matching methods can be used according to the characteristics of different cloud layers and cloud blocks in different geographical areas, which is beneficial to promotion in multiple regions; secondly, the main function of the ground-based cloud image device is to monitor the sky. The present invention adds a new cloud height measurement function on its basis, which can better utilize the existing ground-based cloud image devices. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 It is a schematic diagram of the device structure of the present invention.

[0024] Figure 2 It is a schematic diagram of the method flow of the present invention.

[0025] Figure 3 Schematic diagram of image segmentation of the present invention.

[0026] Figure 4 It is a schematic diagram of the conversion between the world coordinate system and the camera coordinate system of the present invention.

[0027] Figure 5 It is a schematic diagram of the principle of calculating cloud height using epipolar geometry according to the present invention. DETAILED DESCRIPTION

[0028] like Figure 1 As shown, a cloud height measurement device based on ground-based cloud images includes: multiple ground-based cloud image acquisition devices and cloud height measurement devices; the n ground-based cloud image acquisition devices acquire ground-based cloud image images from the ground in real time, and obtain the focal lengths corresponding to different ground-based cloud images and the attitude angles of different ground-based cloud images in real time; the cloud height measurement device processes the ground-based cloud images acquired by the multiple ground-based cloud image acquisition devices to obtain the height information of the target cloud block.

[0029] The cloud height measurement equipment includes a ground-based cloud map image distortion processing module, which is used to perform distortion processing on two or more cloud map images collected at the current sampling moment; a ground-based cloud map image matching module, which is used to match the same target cloud block in two or more ground-based cloud maps to obtain the positions of the edge pixels and the geometric center pixels of the cloud block; a ground-based cloud map coordinate system conversion module, which uses matrix conversion to perform coordinate system conversion according to the focal lengths of different base cloud maps and the attitude angles of the ground-based cloud map acquisition device. The coordinate system conversion is used to convert the pixel positions of the target cloud block pixels via the pixel coordinate system to the image coordinate system, and then from the image coordinate system to the camera coordinate system, and finally from the camera coordinate system to the world coordinate system; a cloud height calculation module, which uses the principle of epipolar geometry in binocular vision and combines the focal length parameters corresponding to multiple cloud maps to calculate the target cloud height for the multiple cloud maps after the coordinate system conversion.

[0030] like Figure 2 and Figure 3 As shown, a cloud height measurement method based on ground-based cloud images includes the following steps:

[0031] Step 1: Input the multi-ground cloud image, the corresponding focal length f, the lens distortion coefficient, the lens attitude angle, and the lens position information into the cloud height calculation module as parameters for subsequent ground-based cloud image fisheye distortion processing, cloud block identification and matching, coordinate system conversion, and epipolar geometry cloud height calculation;

[0032] Step 2: The ground-based cloud image distortion processing module performs distortion correction on the ground-based cloud image images of the same size and overlapped with each other collected by multiple base cloud image collection devices;

[0033] Step 3: The ground-based cloud image matching module identifies and matches the target cloud blocks in the two ground-based cloud images;

[0034] Step 4: The ground-based cloud map coordinate system conversion module converts the camera coordinate system and the world coordinate system according to specific needs;

[0035] Step 5: The cloud height calculation module calculates the target cloud height using the epipolar geometry principle.

[0036] In step 2, the ground-based cloud image distortion processing module measures the lens distortion coefficient of the ground-based cloud image in advance by a chessboard measurement method, and then corrects the image by an existing dot matrix calibration method, and only corrects the area of ​​interest or the required area.

[0037] In step 3, the ground-based cloud image matching module identifies and matches the target cloud blocks in the two ground-based cloud images, specifically including the following steps:

[0038] Step 31, dividing the ground-based cloud image into multiple parts including clouds, sky and sun;

[0039] Step 32: Identify the same cloud block in the two ground-based cloud images through ORB key point matching, and record the pixel coordinates (X1, Y1) and (X2, Y2) of the cloud block in the two images.

[0040] Step 33: If the same cloud block cannot be identified in step 32, the same cloud block in the two ground-based cloud images is identified through the relationship between the target cloud block and the sun in the two ground-based cloud images, and the pixel coordinates (X1, Y1) and (X2, Y2) of the cloud block in the two images are recorded.

[0041] like Figure 4 As shown, in step 4, the world coordinate system is the absolute coordinate system of the real three-dimensional space, and the world coordinate system is converted to the camera coordinate system through the rigid body transformation of rotation and translation; the rigid body transformation only changes the spatial position (translation) and orientation (rotation) of the object, but does not change its shape, which is described by two variables: the orthogonal unit rotation matrix R, the three-dimensional translation vector t, and the origin of the world coordinate system is moved to the origin of the camera coordinate system after translation; the rotation has three degrees of freedom, that is, rotation around three coordinate axes, and the rotation matrices Rx, Ry, and Rz in three directions can be obtained according to the rotation angle, and the rotation matrix is ​​their product: R = Rx*Ry*Rz; different images should be converted to the corresponding ground-based cloud map device coordinate system.

[0042] like Figure 5 As shown in step 5, the cloud height calculation module calculates the target cloud height by using the epipolar geometry principle. Points CL and CR represent two ground-based cloud image acquisition devices at different locations. The world coordinate system OXYZ is precisely transformed and coincides with the camera coordinate system of the CL ground-based cloud image acquisition device. Points PL and PR represent the imaging points of the target cloud block P in the two cloud images. The line segment b is the distance between the two ground-based cloud image devices. The triangle PC L CR is similar to triangle PPLPR, and according to the principle of similar triangles, Then the cloud height is derived

[0043] The present invention uses two ground-based cloud image acquisition devices with the same parameters but in different locations to replace human eyes, collect images of the same cloud target, and calculate the distance information of the target through the difference in the location of the cloud image in the ground-based cloud images collected by two different ground-based cloud image acquisition devices. Compared with expensive ceilometers, ground-based cloud image devices have the advantages of low cost and wide-scale deployment and promotion; compared with the incomplete controllability and unstable nature of measuring cloud height by meteorological balloons, the present invention has strong controllability and stability based on ground-based cloud images, and can monitor and measure sky cloud information at any time.

Claims

1. A cloud height measuring device based on ground-based cloud images, characterized in that: include: Multiple ground-based cloud image acquisition equipment and cloud height measurement equipment; n ground-based cloud map acquisition devices collect ground-based cloud map images from the ground in real time, and obtain the focal lengths and attitude angles corresponding to different ground-based cloud maps in real time. The cloud height measurement device processes the ground-based cloud maps collected by multiple ground-based cloud map acquisition devices to obtain the height information of the target cloud block.

2. The cloud height measuring device based on ground-based cloud images according to claim 1, characterized in that: The cloud height measurement equipment includes a ground-based cloud image distortion processing module, which is used to perform distortion processing on two or more cloud image images collected at the current sampling moment; a ground-based cloud image image matching module, which is used to match the same target cloud block in two or more ground-based cloud images to obtain the positions of the edge pixels and the geometric center pixels of the cloud block; a ground-based cloud image coordinate system conversion module, which uses matrix conversion to perform coordinate system conversion according to the focal lengths of different base cloud images and the attitude angles of the ground-based cloud image acquisition device. The coordinate system conversion is used to convert the pixel positions of the target cloud block pixels into the image coordinate system via the pixel coordinate system, and then from the image coordinate system to the camera coordinate system, and finally from the camera coordinate system to the world coordinate system; The cloud height calculation module calculates the target cloud height of multiple cloud images after the coordinate system is converted by combining the epipolar geometry principle in binocular vision with the focal length parameters corresponding to the multiple cloud images.

3. A measurement method applied to the cloud height measurement device based on ground-based cloud images as claimed in claim 1, characterized in that: The steps include: Step 1: Input the multi-ground cloud image, the corresponding focal length f, the lens distortion coefficient, the lens attitude angle, and the lens position information into the cloud height calculation module as parameters for subsequent ground-based cloud image fisheye distortion processing, cloud block identification and matching, coordinate system conversion, and epipolar geometry cloud height calculation; Step 2: The ground-based cloud image distortion processing module performs distortion correction on the ground-based cloud image images of the same size and overlapped with each other collected by multiple base cloud image collection devices; Step 3: The ground-based cloud image matching module identifies and matches the target cloud blocks in the two ground-based cloud images; Step 4: The ground-based cloud map coordinate system conversion module converts the camera coordinate system and the world coordinate system according to specific needs; Step 5: The cloud height calculation module calculates the target cloud height using the epipolar geometry principle.

4. The cloud height measurement method based on ground-based cloud images according to claim 3, characterized in that: In step 2, the ground-based cloud image distortion processing module measures the lens distortion coefficient of the ground-based cloud image in advance by a chessboard measurement method, and then corrects the image by an existing dot matrix calibration method, and only corrects the area of ​​interest or the required area.

5. The cloud height measurement method based on ground-based cloud images according to claim 4, characterized in that: The lens distortion coefficient is measured by the chessboard calibration method. After the lens distortion coefficient of each ground-based cloud image acquisition device is obtained by the chessboard calibration method, the corrected image is obtained by the image distortion correction method.

6. The cloud height measurement method based on ground-based cloud images according to claim 3, characterized in that: In step 3, the ground-based cloud image matching module identifies and matches the target cloud blocks in the two ground-based cloud images, specifically including the following steps: Step 31, segmenting the ground-based cloud image into multiple parts including cloud, sky and sun by using the red-to-blue ratio method; Step 32: Identify the same cloud block in the two ground-based cloud images through ORB key point matching, and record the pixel coordinates (X1, Y1) and (X2, Y2) of the cloud block in the two images. Step 33: If the same cloud block cannot be identified in step 32, the same cloud block in the two ground-based cloud images is identified through the relationship between the target cloud block and the sun in the two ground-based cloud images, and the pixel coordinates (X1, Y1) and (X2, Y2) of the cloud block in the two images are recorded.

7. The cloud height measurement method based on ground-based cloud images according to claim 3, characterized in that: In step 4, the world coordinate system is the absolute coordinate system of the real three-dimensional space. The world coordinate system is converted into the camera coordinate system through rigid body transformation of rotation and translation; the orthogonal unit rotation matrix R, the three-dimensional translation vector t, the origin of the world coordinate system is moved to the origin of the camera coordinate system after translation; the rotation has three degrees of freedom, that is, rotation around the three coordinate axes, and the rotation matrices Rx, Ry, and Rz in the three directions can be obtained according to the rotation angle, and the rotation matrix is ​​their product: R=Rx*Ry*Rz; different images should be converted to the corresponding ground-based cloud map device coordinate system.

8. The cloud height measurement method based on ground-based cloud images according to claim 3, characterized in that: In step 5, the cloud height calculation module calculates the target cloud height by using the principle of epipolar geometry. Points CL and CR represent two ground-based cloud image acquisition devices at different locations. The world coordinate system OXYZ is precisely transformed and coincides with the camera coordinate system of the CL ground-based cloud image acquisition device. Points PL and PR represent the imaging points of the target cloud block P in the two cloud images. Line segment b is the distance between the two ground-based cloud image devices. Triangle PC L C R and triangle PP L P R Similar, according to the principle of similar triangles Then the cloud height is derived

Images