Multi-camera vertical direction coordinate calibration method and device and electronic equipment

Abstract

The invention provides a multi-camera vertical direction coordinate calibration method and device and electronic equipment, multiple cameras comprise a first camera and at least one second camera, and an electronic compass is arranged in the first camera. The method comprises the following steps: in a vertical plane, controlling a first camera to face a preset direction and collect a first image, and controlling a second camera to rotate from an initial position along a set direction and collect a second image; lenses of the first camera and the second camera face the same direction in the horizontal direction; obtaining the rotation angle of the second camera and the number of feature points in the overlapping area of the second image and the first image at each rotation angle; according to the corresponding relation between the rotation angles and the number of the feature points in the overlapped area under each rotation angle, obtaining an overlapped feature point change curve; the vertical direction coordinates of the second camera are calibrated according to the curve and the vertical direction coordinates of the first camera, the hardware cost of multiple cameras is reduced, interference of external factors is eliminated, and the accuracy of calibration results is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of camera, in particular to a multi-camera vertical direction coordinate calibration method and device and electronic equipment. BACKGROUND

[0002] The multi-camera product includes a panoramic camera and a detail camera. Generally, the panoramic camera is used to capture the overall picture of the outside world, and the detail camera is used to track the target. In actual application, the multi-camera needs to be calibrated in coordinates first, so that the pictures taken by the panoramic camera and the detail camera in the multi-camera can correspond, thereby realizing accurate shooting and tracking of the target.

[0003] At present, for the coordinates in the vertical direction, the multi-camera coordinate calibration is mainly realized by adding an electronic compass in each camera in the multi-camera. However, since the electronic compass has a high cost, the current calibration method increases the hardware cost of the multi-camera. SUMMARY

[0004] The present application provides a multi-camera vertical direction coordinate calibration method and device and electronic equipment, which solves the defect that the multi-camera needs to install an electronic compass in each camera in the prior art, realizes that only one electronic compass needs to be reserved in the multi-camera, and reduces the hardware cost accuracy of the multi-camera.

[0005] The present application provides a multi-camera vertical direction coordinate calibration method, the multi-camera includes a first camera and at least one second camera, and the first camera is provided with an electronic compass. The method comprises the following steps: in a vertical plane, controlling the first camera to face a preset direction and collecting a first image, and controlling the second camera to rotate from an initial position along a set direction and collecting a second image; wherein the lenses of the first camera and the second camera are consistent in the horizontal direction; obtaining the rotation angle of the second camera, and obtaining the number of feature points in the overlapping area of the second image and the first image under each rotation angle; obtaining the change curve of the overlapping feature points according to the corresponding relationship between the rotation angle and the number of feature points in the overlapping area under each rotation angle; and calibrating the vertical direction coordinates of the second camera according to the change curve of the overlapping feature points and the vertical direction coordinates of the first camera determined based on the electronic compass.

[0006] According to the multi-camera vertical direction coordinate calibration method provided by the application, the field angles of the lenses of the first camera and the second camera are different, the number of feature points in the overlapping area of the second image and the first image at each rotation angle is obtained, and the method comprises the following steps: determining a large field angle image and a small field angle image in the second image and the first image according to the field angles of the lenses of the first camera and the second camera; wherein the field angle of the camera corresponding to the large field angle image is greater than the field angle of the camera corresponding to the small field angle image; determining an effective comparison area in the large field angle image according to the small field angle image; and obtaining the number of feature points in the overlapping area of the effective comparison area in the large field angle image and the small field angle image at each rotation angle.

[0007] According to the multi-camera vertical direction coordinate calibration method provided by the application, the effective comparison area in the large field angle image is determined according to the small field angle image, and the method comprises the following steps: in the large field angle image, an area with the center of the large field angle image as the center and the imaging circle radius of the small field angle image as the radius is determined as the effective comparison area.

[0008] According to the multi-camera vertical direction coordinate calibration method provided by the application, the vertical direction coordinate of the second camera is calibrated according to the overlapping feature point change curve and the vertical direction coordinate of the first camera determined based on the electronic compass, and the method comprises the following steps: determining whether there is a first rotation angle in a period of the overlapping feature point change curve according to the number of feature points in the overlapping area at each point on the overlapping feature point change curve; wherein the number of feature points in the overlapping area corresponding to the first rotation angle is greater than the number of feature points in the overlapping area corresponding to other rotation angles in the period of the first rotation angle; in the case that the first rotation angle exists, the second camera is rotated by the first rotation angle from the initial position along a set direction; and the vertical direction coordinate of the second camera is adjusted to be consistent with the first camera according to the coordinate of the first camera in the vertical direction.

[0009] According to the multi-camera vertical direction coordinate calibration method provided by the application, the vertical direction coordinate of the second camera is calibrated according to the coincidence feature point change curve and the vertical direction coordinate of the first camera determined based on the electronic compass, and the method further comprises the following steps: in the absence of the first rotation angle, a plurality of target inflection points are determined within one period of the coincidence feature point change curve; wherein the plurality of target inflection points comprise a first target inflection point at which the number of feature points in the overlapping area on the coincidence feature point change curve increases as the rotation angle increases and is greater than a set value, and a second target inflection point at which the number of feature points in the overlapping area on the coincidence feature point change curve decreases as the rotation angle decreases and is less than the set value; a second rotation angle is calculated according to the reference rotation angle corresponding to each target inflection point in the plurality of target inflection points on the coincidence feature point change curve; the second camera is rotated by the second rotation angle in the set direction from the initial position; and the vertical direction coordinate of the second camera is adjusted to be consistent with that of the first camera according to the coordinate of the first camera in the vertical direction.

[0010] According to the multi-camera vertical direction coordinate calibration method provided by the application, the vertical direction coordinate of the second camera is calibrated according to the coincidence feature point change curve and the vertical direction coordinate of the first camera determined based on the electronic compass, and the method further comprises the following steps: in the absence of the first rotation angle, a plurality of target inflection points are determined within one period of the coincidence feature point change curve; wherein the plurality of target inflection points comprise a first target inflection point at which the number of feature points in the overlapping area on the coincidence feature point change curve increases as the rotation angle increases and is greater than a set value, and a second target inflection point at which the number of feature points in the overlapping area on the coincidence feature point change curve decreases as the rotation angle decreases and is less than the set value; a second rotation angle is calculated according to the reference rotation angle corresponding to each target inflection point in the plurality of target inflection points on the coincidence feature point change curve; the second camera is rotated by the second rotation angle in the set direction from the initial position; and the vertical direction coordinate of the second camera is adjusted to be consistent with that of the first camera according to the coordinate of the first camera in the vertical direction.

[0011] The application further provides a multi-camera vertical direction coordinate calibration device, wherein the multi-camera comprises a first camera and at least one second camera, and the first camera is provided with an electronic compass, and the device comprises a control module, an acquisition module, a processing module and a calibration module.

[0012] The control module is used for controlling the first camera to face a preset direction and collect a first image, and controlling the second camera to rotate from an initial position in a set direction and collect a second image in a vertical plane; wherein the lenses of the first camera and the second camera face in the same direction in the horizontal direction; The acquisition module is used for acquiring the rotation angle of the second camera, and acquiring the number of feature points in the overlapping area between the second image and the first image under each rotation angle; The processing module is used for obtaining a coincidence feature point change curve according to the corresponding relationship between the rotation angle and the number of feature points in the overlapping area under each rotation angle; The calibration module is used for calibrating the vertical direction coordinate of the second camera according to the coincidence feature point change curve and the vertical direction coordinate of the first camera determined based on the electronic compass.

[0013] The application further provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the multi-camera vertical direction coordinate calibration method when executing the program.

[0014] The application further provides a computer readable storage medium, which stores a computer program, and the computer program implements the multi-camera vertical direction coordinate calibration method when executed by a processor.

[0015] The application further provides a computer program product, which comprises a computer program, and the computer program implements the multi-camera vertical direction coordinate calibration method when executed by a processor.

[0016] The application provides a multi-camera vertical direction coordinate calibration method and device, and an electronic device, wherein the multi-camera comprises a first camera and at least one second camera, and the first camera is provided with an electronic compass, the application controls the first camera to face a preset direction and collect a first image in a vertical plane, and controls the second camera to rotate from an initial position along a set direction and collect a second image, wherein the lenses of the first camera and the second camera face the same direction in a horizontal direction, the rotation angle of the second camera is obtained, and the number of feature points in the overlapping area of the second image and the first image under each rotation angle is obtained, the corresponding relationship between the rotation angle and the number of feature points in the overlapping area under each rotation angle is obtained to obtain a change curve of the overlapping feature points, and the vertical direction coordinate of the second camera is calibrated according to the change curve of the overlapping feature points and the vertical direction coordinate of the first camera determined based on the electronic compass. Therefore, the application can calibrate the vertical direction coordinate of the multi-camera when only one electronic compass is arranged in the multi-camera, solve the defects that the multi-camera needs to be provided with an electronic compass in each camera in the prior art, the hardware cost of the multi-camera is increased, and the multi-camera is easily disturbed by external factors, so that only one electronic compass needs to be reserved in the multi-camera, the hardware cost of the multi-camera is reduced, the external factor disturbance is effectively excluded, and the accuracy of the coordinate calibration result is improved. BRIEF DESCRIPTION OF DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the present application or the prior art, the following will briefly introduce the drawings needed in the embodiments or prior art description. Obviously, the drawings in the following description are some embodiments of the present application, and those skilled in the art can obtain other drawings according to these drawings without creative labor.

[0018] Figure 1 is one of the flowcharts of the multi-camera vertical direction coordinate calibration method provided by the embodiments of the present application.

[0019] Figure 2 is a schematic diagram of setting orientation in the multi-camera vertical direction coordinate calibration method provided by the embodiment of the present application.

[0020] Figure 3 is a schematic diagram of the first image being a large field of view angle image and the second image being a small field of view angle image in the multi-camera vertical direction coordinate calibration method provided by the embodiment of the present application.

[0021] Figure 4 is a schematic diagram of the first image being a small field of view angle image and the second image being a large field of view angle image in the multi-camera vertical direction coordinate calibration method provided by the embodiment of the present application.

[0022] Figure 5 is a second flowchart of the multi-camera vertical direction coordinate calibration method provided by the embodiment of the present application.

[0023] Figure 6 is a schematic diagram of the effective comparison area in the multi-camera vertical direction coordinate calibration method provided by the embodiment of the present application.

[0024] Figure 7 is a schematic diagram of the position relationship between the first image and the second image in various cases corresponding to the relative position change of the first camera and the second camera in the vertical plane in the multi-camera vertical direction coordinate calibration method provided by the embodiment of the present application.

[0025] Figure 8 is a first schematic diagram of the change curve of the coincident feature points in the multi-camera vertical direction coordinate calibration method provided by the embodiment of the present application.

[0026] Figure 9 is a second schematic diagram of the change curve of the coincident feature points in the multi-camera vertical direction coordinate calibration method provided by the embodiment of the present application.

[0027] Figure 10 is a third flowchart of the multi-camera vertical direction coordinate calibration method provided by the embodiment of the present application.

[0028] Figure 11 is a fourth flowchart of the multi-camera vertical direction coordinate calibration method provided by the embodiment of the present application.

[0029] Figure 12 is a schematic diagram of the position relationship change between the first image and the second image in the multi-camera vertical direction coordinate calibration method provided by the embodiment of the present application.

[0030] Figure 13 is a structural schematic diagram of the multi-camera vertical direction coordinate calibration device provided by the embodiment of the present application.

[0031] Figure 14is a structural schematic diagram of an electronic device provided by the present application. DETAILED DESCRIPTION

[0032] In order to make the objects, technical solutions and advantages of the present application clearer, the technical solutions in the present application will be described clearly and completely below with reference to the drawings in the present application. Obviously, the described embodiments are only some of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present application.

[0033] The multi-camera vertical direction coordinate calibration method for describing the embodiments of the present application will be described below. Figures 1-12 The multi-camera of the present application includes a first camera and at least one second camera, and an electronic compass is arranged in the first camera.

[0034] Figure 1 is one of flowcharts of the multi-camera vertical direction coordinate calibration method provided by the present application. As shown in Figure 1 the multi-camera vertical direction coordinate calibration method of the present application includes the following S110-S140.

[0035] S110: in a vertical plane, the first camera is controlled to face a preset direction and collect a first image, and the second camera is controlled to rotate from an initial position along a set direction and collect a second image; wherein the lenses of the first camera and the second camera face the same in the horizontal direction.

[0036] The lenses of the first camera and the second camera can be first adjusted to face the same in the horizontal direction, for example, the lenses of the first camera and the second camera can be adjusted to face the north direction in the horizontal direction. In a specific implementation, the gimbals in the first camera and the second camera can be adjusted to rotate in the horizontal direction, so that the lenses of the first camera and the second camera face the same in the horizontal direction.

[0037] In the vertical plane, the first camera is controlled to face a preset direction and collect a first image.

[0038] The posture of the first camera is unchanged in the process of collecting the first image. The facing direction of the first camera in the vertical plane can be any facing direction.

[0039] For example, the facing direction of the first camera in the vertical plane can be facing the ground, as shown in Figure 2 the lens of the first camera 210 faces the ground in the vertical plane; for another example, the facing direction of the first camera in the vertical plane can be facing the indoor ceiling (or the sky), as shown in Figure 2As shown in the middle part of FIG. 2A, the lens of the first camera 210 is directed to the indoor ceiling in the vertical plane; as another example, the direction of the first camera in the vertical plane can be directed to the horizontal direction, as shown in the middle part of FIG. 2B. Figure 2 As shown in the middle part of FIG. 2C, the lens of the first camera 210 is directed to the horizontal direction in the vertical plane.

[0040] The second camera rotates from the initial position along a set direction, which can be, for example, the counterclockwise direction in the vertical plane or the clockwise direction in the vertical plane.

[0041] The speed of the rotation along the set direction can be uniform, and the speed value can be set by those skilled in the art according to actual conditions, and the present application does not limit this.

[0042] The speed of the rotation along the set direction can also not be uniform, for example, it can be accelerated in the case where the first image and the second image have no overlapping area, and it becomes uniform in the case where the first image and the second image have overlapping areas; as another example, it can be accelerated or uniform in the case where the first image and the second image have no overlapping area, and it becomes gradually decelerated in the case where the first image and the second image have overlapping areas, and so on.

[0043] S120: Obtain the rotation angle of the second camera, and obtain the number of feature points in the overlapping area of the second image and the first image at each rotation angle.

[0044] Since the direction of the first camera in the vertical direction does not change, the first image collected by the first camera at different times is the same. The second camera rotates along the set direction, and the second image collected by the second camera at different times is different.

[0045] The first image is the imaging result of the lens of the first camera, and its shape is an imaging circle (i.e., a circular shape). Similarly, the second image is the imaging result of the lens of the second camera, and its shape is an imaging circle.

[0046] In some embodiments, the field angles of the lenses of the first camera and the second camera are the same, in which case the imaging circle corresponding to the imaging result of the first camera and the imaging circle corresponding to the imaging result of the second camera are the same in size, and they are two imaging circles of the same size. When obtaining the number of feature points in the overlapping area of the second image and the first image at each rotation angle, the number of feature points in the overlapping area of the second image and the first image can be directly obtained.

[0047] In some embodiments, the field angles of the lenses of the first camera and the second camera are different, in which case the imaging circle corresponding to the imaging result of the first camera and the imaging circle corresponding to the imaging result of the second camera are different in size.

[0048] The field of view of the lens of the first camera and the field of view of the lens of the second camera being different can include two cases: the field of view of the lens of the first camera being greater than the field of view of the lens of the second camera, and the field of view of the lens of the first camera being less than the field of view of the lens of the second camera.

[0049] In the case where the field of view of the lens of the first camera is greater than the field of view of the lens of the second camera, as shown in the a part and the b part of FIG. 6, the radius of the imaging circle corresponding to the imaging result of the first camera (the a part of FIG. 6) is greater than the radius of the imaging circle corresponding to the imaging result of the second camera (the b part of FIG. 6). Figure 3 Figure 3 In the case where the field of view of the lens of the first camera is greater than the field of view of the lens of the second camera, as shown in the a part and the b part of FIG. 6, the radius of the imaging circle corresponding to the imaging result of the first camera (the a part of FIG. 6) is greater than the radius of the imaging circle corresponding to the imaging result of the second camera (the b part of FIG. 6). Figure 3

[0050] In the case where the field of view of the lens of the first camera is less than the field of view of the lens of the second camera, as shown in the a part and the b part of FIG. 7, the radius of the imaging circle corresponding to the imaging result of the first camera (the a part of FIG. 7) is less than the radius of the imaging circle corresponding to the imaging result of the second camera (the b part of FIG. 7). Figure 4 Figure 4 In the case where the field of view of the lens of the first camera is less than the field of view of the lens of the second camera, as shown in the a part and the b part of FIG. 7, the radius of the imaging circle corresponding to the imaging result of the first camera (the a part of FIG. 7) is less than the radius of the imaging circle corresponding to the imaging result of the second camera (the b part of FIG. 7). Figure 4

[0051] In the case where the field of view of the lens of the first camera and the field of view of the lens of the second camera are different, when the number of feature points in the overlapping area of the second image and the first image at each rotation angle is acquired, as shown in FIG. 8, the following S510-S530 can be included. Figure 5

[0052] S510: According to the field of view of the lens of the first camera and the field of view of the lens of the second camera, a large field of view image and a small field of view image are determined in the second image and the first image; wherein the field of view of the corresponding camera of the large field of view image is greater than the field of view of the corresponding camera of the small field of view image.

[0053] The radius of the imaging circle of the large field of view image is greater than the radius of the imaging circle of the small field of view image. Therefore, among the imaging circle corresponding to the imaging result of the first camera and the imaging circle corresponding to the imaging result of the second camera, the image corresponding to the imaging circle with the larger radius is taken as the large field of view image, and the image corresponding to the imaging circle with the smaller radius is taken as the small field of view image.

[0054] In the case where the field of view of the lens of the first camera is greater than the field of view of the lens of the second camera, referring to the a part and the b part of FIG. 6, the first image collected by the first camera is determined as the large field of view image, and the second image collected by the second camera is determined as the small field of view image. Figure 3

[0055] In the case where the field of view of the lens of the first camera is less than the field of view of the lens of the second camera, referring to the a part and the b part of FIG. 7, the first image collected by the first camera is determined as the large field of view image, and the second image collected by the second camera is determined as the small field of view image. Figure 4 ​​​​​​In the part a and the part b, the first image captured by the first camera is determined as a small field angle image, and the second image captured by the second camera is determined as a large field angle image.

[0056] S520: determining an effective matching area in the large field angle image according to the small field angle image.

[0057] In the large field angle image, an area with the center of the large field angle image as the center and the imaging circle radius of the small field angle image as the radius is determined as the effective matching area.

[0058] For example, as shown in Figure 6 , the imaging circle corresponding to the large field angle image is imaging circle A, and the imaging circle corresponding to the small field angle image is imaging circle B. In the large field angle image, an area with the center of the large field angle image as the center and the imaging circle radius of the small field angle image as the radius is determined as the effective matching area (as shown in the shadow part of Figure 6 ).

[0059] S530: obtaining the number of feature points in the overlapping area where the effective matching area in the large field angle image and the small field angle image overlap under each rotation angle.

[0060] In the case where the first image is a large field angle image and the second image is a small field angle image, the number of feature points in the overlapping area where the second image and the effective matching area in the first image overlap under each rotation angle is obtained.

[0061] In the case where the first image is a small field angle image and the second image is a large field angle image, the number of feature points in the overlapping area where the first image and the effective matching area in the second image overlap under each rotation angle is obtained.

[0062] Next, combined with Figure 7 , the process of S120 is described taking the case where the field angles of the lenses of the first camera and the second camera are the same as an example.

[0063] Referring to Figure 7 , the part a, the part b, the part c and the part d show that the set orientation of the first camera is that the lens faces the ground, and during the rotation of the second camera, the positional relationship between the second image captured by the lens of the second camera and the first image captured by the lens of the first camera includes the following three cases: apart, tangent and intersect.

[0064] In the case where the first image and the second image are apart, as shown in the part a of Figure 7 , the lens of the first camera 710 faces the ground, the initial position of the lens of the second camera 720 faces the ceiling, and the first image C and the second image D do not overlap, i.e., there is no overlapping area.

[0065] Next, asFigure 7 As shown in the middle b part, the first camera 710 is still lens down, and the second camera 720 is from the first position shown in the middle b part to the current second position along the counterclockwise direction at a constant speed. In the vertical plane, the posture of the second camera 720 at the second position is more close to the posture of the first camera 710 than the posture of the second camera 720 at the first position. The first image C and the second image D appear an overlapping area (as shown by the shadow in the middle c part), and the overlapping area becomes larger and larger with the continuous rotation of the second camera 720. Figure 7 As shown in the middle a part, the first camera 710 is still lens down, and the second camera 720 is from the initial position shown in the middle a part to the current first position along the counterclockwise direction at a constant speed. At this time, the first image C and the second image D appear an intersecting intersection. As can be seen, with the rotation of the second camera 720, the first image C and the second image D change from apart to tangent, and the intersection changes from 0 to 1.

[0066] Next, as shown in the middle c part, the first camera 710 is still lens down, and the second camera 720 is from the first position shown in the middle b part to the current second position along the counterclockwise direction at a constant speed. In the vertical plane, the posture of the second camera 720 at the second position is more close to the posture of the first camera 710 than the posture of the second camera 720 at the first position. The first image C and the second image D appear an overlapping area (as shown by the shadow in the middle c part), and the overlapping area becomes larger and larger with the continuous rotation of the second camera 720. Figure 7 As shown in the middle a part, the first camera 710 is still lens down, and the second camera 720 is from the initial position shown in the middle a part to the current first position along the counterclockwise direction at a constant speed. At this time, the first image C and the second image D appear an intersecting intersection. As can be seen, with the rotation of the second camera 720, the first image C and the second image D change from apart to tangent, and the intersection changes from 0 to 1. Figure 7 As shown in the middle b part, the first camera 710 is still lens down, and the second camera 720 is from the first position shown in the middle b part to the current second position along the counterclockwise direction at a constant speed. In the vertical plane, the posture of the second camera 720 at the second position is more close to the posture of the first camera 710 than the posture of the second camera 720 at the first position. The first image C and the second image D appear an overlapping area (as shown by the shadow in the middle c part), and the overlapping area becomes larger and larger with the continuous rotation of the second camera 720. Figure 7 As shown in the middle a part, the first camera 710 is still lens down, and the second camera 720 is from the initial position shown in the middle a part to the current first position along the counterclockwise direction at a constant speed. At this time, the first image C and the second image D appear an intersecting intersection. As can be seen, with the rotation of the second camera 720, the first image C and the second image D change from apart to tangent, and the intersection changes from 0 to 1.

[0067] Finally, as shown in the middle d part, the first camera 710 is still lens down, and the second camera 720 is also from the second position to the second camera 720 is also lens down along the counterclockwise direction at a constant speed. In this case, in the vertical plane, the posture of the second camera 720 is consistent with the posture of the first camera 710, and the overlapping area (as shown by the shadow in the middle d part) of the first image and the second image is the largest. Figure 7 As shown in the middle a part, the first camera 710 is still lens down, and the second camera 720 is from the initial position shown in the middle a part to the current first position along the counterclockwise direction at a constant speed. At this time, the first image C and the second image D appear an intersecting intersection. As can be seen, with the rotation of the second camera 720, the first image C and the second image D change from apart to tangent, and the intersection changes from 0 to 1. Figure 8 As shown in the middle a part, the first camera 710 is still lens down, and the second camera 720 is from the initial position shown in the middle a part to the current first position along the counterclockwise direction at a constant speed. At this time, the first image C and the second image D appear an intersecting intersection. As can be seen, with the rotation of the second camera 720, the first image C and the second image D change from apart to tangent, and the intersection changes from 0 to 1.

[0068] In the art, the extraction of image feature vectors can be used for analyzing image similarity, searching for images, etc. For two adjacent cameras, the similarity between the non-overlapping areas of the fields of view is often lower than the similarity between the overlapping areas. When the overlapping area is small, the number of feature points in the overlapping area is also small. With the increase of the overlapping area, the number of feature points in the overlapping area also increases. Therefore, by knowing the change of the number of feature points in the overlapping area, the change of the size of the overlapping area can be known, and when the number of feature points in the overlapping area is the largest, the overlapping area is the largest.

[0069] S130: Obtain the overlapping feature point change curve according to the corresponding relationship between the rotation angle and the number of feature points in the overlapping area at each rotation angle.

[0070] Taking the rotation angle as the independent variable and the number of feature points in the overlapping area at each rotation angle as the dependent variable, a fitting curve of the relationship between the rotation angle and the number of feature points in the overlapping area is generated as the overlapping feature point change curve.

[0071] For example, asFigure 8 As shown in FIG. 6, with the rotation angle as the independent variable and the number of feature points in the overlapping region at each rotation angle as the dependent variable, a curve of the number of overlapping feature points is obtained.

[0072] S140: According to the curve of the number of overlapping feature points and the vertical direction coordinate of the first camera determined based on the electronic compass, the vertical direction coordinate of the second camera is calibrated.

[0073] Firstly, the rotation angle corresponding to the maximum number of feature points in the overlapping region in the curve of the number of overlapping feature points can be determined.

[0074] Specifically, in the curve of the number of overlapping feature points, the maximum number of feature points in the overlapping region in one period can have only one (as shown in FIG. 6); or, the maximum number of feature points in the overlapping region in one period can have multiple (as shown in FIG. 7), for example, the maximum number of feature points in the overlapping region is not clear, so that the maximum number of feature points in the current overlapping region corresponds to multiple rotation angles. Figure 9 Figure 10 In some embodiments, in the case that the maximum number of feature points in the overlapping region in one period has only one, as shown in FIG. 6, the execution process of S140 can include S1010-S1030.

[0075] In some embodiments, in the case that the maximum number of feature points in the overlapping region in one period has only one, as shown in FIG. 6, the execution process of S140 can include S1010-S1030. Figure 8

[0076] S1010: According to the number of feature points in the overlapping region at each point in the curve of the number of overlapping feature points, it is determined whether there is a first rotation angle in one period of the curve of the number of overlapping feature points; wherein the number of feature points in the overlapping region corresponding to the first rotation angle is greater than the number of feature points in the overlapping region corresponding to other rotation angles in the period of the first rotation angle.

[0077] The first rotation angle exists in the curve of the number of overlapping feature points, that is, in one period, there is only one maximum number of feature points in the overlapping region in the curve of the number of overlapping feature points. For example Figure 7 Three periods are shown in FIG. 6, and each period has a maximum number of feature points in the overlapping region, in which case it is determined that the first rotation angle exists.

[0078] The way to determine whether there is a first rotation angle in one period of the curve of the number of overlapping feature points can be set by those skilled in the art according to the actual situation, and the present application does not limit it.

[0079] S1020: In the case that the first rotation angle exists, the second camera is rotated by the first rotation angle from the initial position along the set direction.

[0080] ​​Since the number of feature points in the coincident region corresponding to the first rotation angle is the largest, it can be determined that the orientation of the second camera lens in the vertical direction is consistent with that of the first camera lens after the second camera lens is rotated from the initial position by the first rotation angle in the set direction (for example Figure 11 As shown in part d of FIG. 13, the second camera lens is controlled to rotate from the initial position by the first rotation angle in the set direction so that the orientation of the second camera lens in the vertical direction is consistent with that of the first camera lens.

[0081] S1030: Adjust the vertical direction coordinate of the second camera to be consistent with that of the first camera according to the coordinate of the first camera in the vertical direction.

[0082] Since the electronic compass is arranged on the first camera, the coordinate of the first camera in the vertical direction can be determined. In this case, the orientation of the second camera lens in the vertical direction has been made consistent with that of the first camera lens in S1020, and thus it can be considered that the coordinate of the second camera in the vertical direction at the current position is consistent with that of the first camera. Therefore, the vertical direction coordinate of the second camera is adjusted to be consistent with that of the first camera, so as to calibrate the vertical direction coordinate of the second camera and unify the coordinates of the first camera and the second camera in the vertical direction.

[0083] For example, the first rotation angle is α, the first camera lens always faces the ground in the vertical plane, and after the first rotation angle is determined and the second camera lens is rotated from the initial position by the first rotation angle in the set direction, it can be determined that the second camera lens also faces the ground in the vertical plane, and the second camera is calibrated by the coordinate of the first camera lens facing the ground in the vertical plane. After that, it can be determined that the second camera lens points to the ceiling (or the sky) after being rotated by β = 180°, and the second camera lens is rotated by β, and then the second camera is calibrated by the coordinate of the first camera lens facing the ceiling in the vertical plane; similarly, the second camera lens points to the horizontal direction after being rotated by γ = (α + β) / 2, and at this time, the second camera lens can be calibrated according to the coordinate of the first camera lens facing the horizontal direction in the vertical direction.

[0084] In some embodiments, the determination result of S1010 is that there is no first rotation angle corresponding to the case that the number of feature points in the coincident region within one period is multiple, in which case, as shown in FIG. 14, S140 can perform S1110-S1140 as follows. Figure 12

[0085] ​S1110: In the absence of a first rotation angle, determine multiple target inflection points within one period of the overlapping feature point change curve; wherein, the multiple target inflection points include: a first target inflection point when the number of feature points in the overlapping area on the overlapping feature point change curve increases with the rotation angle and is greater than a set value, and a second target inflection point when the number of feature points in the overlapping area on the overlapping feature point change curve decreases with the rotation angle and is less than a set value.

[0086] For a detailed explanation of the first rotation angle, please refer to the corresponding description in S1010, which will not be repeated here.

[0087] There is no first rotation angle, meaning that the number of feature points in the overlapping region within one cycle of the curve of overlapping feature points has multiple maximum values.

[0088] In some embodiments, such as Figure 12 As shown, the positional changes of the first image C and the second image D will sequentially go through the following stages: separation, from separation to tangency, from tangency to intersection, the intersection with the largest intersection area, from intersection to tangency, and from tangency to separation.

[0089] See Figure 12 The first image C and the second image D start to overlap from the first time they are tangent. The overlapping area gradually increases, and the number of feature points in the overlapping area increases accordingly until the maximum number of feature points in the overlapping area is reached. Then the first image C and the second image D change from intersecting to tangent. The overlapping area gradually decreases, and the number of feature points in the overlapping area decreases accordingly until the first image C and the second image D finally change from tangent to disjoint.

[0090] See Figure 12 The first image C and the second image D change from being separate to being tangent, which is the first switching state. The first image C and the second image D change from being tangent to being separate, which is the second switching state. The second switching state is located after the first switching state.

[0091] See Figure 9 Based on the changes in the size of the overlapping region, it can be seen that the maximum number of feature points within the overlapping region occurs precisely in the middle of the transition from the first switching state to the second switching state. Corresponding to the curve of the overlapping feature point change, the first target inflection point appears when the first switching state occurs, and the second target inflection point appears when the second switching state occurs. The rotation angle corresponding to the maximum number of feature points within the overlapping region should be the rotation angle corresponding to the midpoint of the line segment between the first and second target inflection points.

[0092] Ideally, see Figure 9 In two adjacent cycles, the number of feature points in the overlapping region on the curve of the overlapping feature points in the previous cycle begins to decrease until the number changes from a positive value to 0, i.e.Figure 9 The position of the rotation angle X1 corresponds to the second target inflection point of the previous period, and the position of the rotation angle X2 corresponds to the first target inflection point of the next period. As can be seen, in the ideal case, the first target inflection point is the inflection point at which the number of feature points in the overlapping region on the coincident feature point change curve changes from a positive value to 0, and the second target inflection point is the inflection point at which the number of feature points in the overlapping region on the coincident feature point change curve changes from 0 to a positive value. Figure 9 At the beginning of the next period mentioned above, the number of feature points in the overlapping region on the coincident feature point change curve does not change and is always 0 as the rotation angle changes. When the rotation angle is X2, the number of feature points in the overlapping region on the coincident feature point change curve is still 0. When the rotation angle exceeds X2, the number of feature points in the overlapping region on the coincident feature point change curve changes from 0 to a positive value.

[0093] The position of the rotation angle X1 corresponds to the second target inflection point of the previous period, and the position of the rotation angle X2 corresponds to the first target inflection point of the next period. As can be seen, in the ideal case, the first target inflection point is the inflection point at which the number of feature points in the overlapping region on the coincident feature point change curve changes from a positive value to 0, and the second target inflection point is the inflection point at which the number of feature points in the overlapping region on the coincident feature point change curve changes from 0 to a positive value.

[0094] For two adjacent periods, referring to Figure 13 , the first target inflection point (the position of X1) of the previous period and the second target inflection point (the position of X2) of the next period should be a smooth straight line in the ideal case. However, in the actual case, even if the two images do not intersect, there may be a small number of identical feature points, which leads to the fact that the second target inflection point of the previous period and the first target inflection point of the next period are not a smooth straight line but a fluctuating curve in the actual obtained coincident feature point change curve. In this case, when calibrating the positions of the first target inflection point and the second target inflection point, if the point at which the number of feature points in the overlapping region on the coincident feature point change curve changes from 0 to a positive value is taken as the calibration point, the result obtained is not accurate.

[0095] Therefore, in the embodiment of the present application, corresponding to the first target inflection point and the second target inflection point in the ideal case, a setting value is set. When the number of feature points in the overlapping region on the coincident feature point change curve changes as the rotation angle increases and is greater than the setting value, it is determined that the position is the first target inflection point. When the number of feature points in the overlapping region on the coincident feature point change curve changes as the rotation angle decreases and is less than the setting value, it is determined that the position is the second target inflection point. In this way, a similar calibration result to the first target inflection point position and the second target inflection point position calibrated in the ideal case can be achieved. The setting value is set by a person skilled in the art according to the actual situation.

[0096] If the number of the feature points in the overlapping region in the overlapping feature point change curve increases with the rotation angle and is greater than a set value at a first target inflection point corresponding to a rotation angle A1, and the number of the feature points in the overlapping region in the overlapping feature point change curve decreases with the rotation angle and is less than a set value at a second target inflection point corresponding to a rotation angle A2.

[0097] S1120: Calculate the second rotation angle according to the reference rotation angle corresponding to each target inflection point in the overlapping feature point change curve.

[0098] The reference rotation angles corresponding to each target inflection point in the overlapping feature point change curve can be summed to obtain a sum result.

[0099] Then, a ratio of the sum result to the number of the target inflection points is obtained, and the ratio is taken as the second rotation angle.

[0100] For example, the reference rotation angles corresponding to the target inflection points in the overlapping feature point change curve include A1 and A2, and the second rotation angle is (A1+A2) / 2.

[0101] S1130: Rotate the second camera from the initial position by the second rotation angle in the set direction.

[0102] S1140: Adjust the vertical direction coordinate of the second camera to be consistent with the first camera according to the coordinate of the first camera in the vertical direction.

[0103] The execution process of S1140 is the same as that of S1030, and specific descriptions can be referred to the corresponding descriptions in S1030, which will not be repeated here.

[0104] The multi-camera vertical coordinate calibration method provided by this invention includes a first camera and at least one second camera. The first camera is equipped with an electronic compass. This invention controls the first camera to face a preset direction and acquire a first image, and controls the second camera to rotate from an initial position along a set direction and acquire a second image within a vertical plane. The lenses of the first and second cameras face the same direction in the horizontal direction. The method obtains the rotation angle of the second camera and the number of feature points in the overlapping area of ​​the second and first images at each rotation angle. Based on the correspondence between the rotation angle and the number of feature points in the overlapping area at each rotation angle, a curve showing the change in overlapping feature points is obtained. The vertical coordinates of the second camera are calibrated based on the curve showing the change in overlapping feature points and the vertical coordinates of the first camera determined by the electronic compass. Therefore, this invention can calibrate the vertical coordinates of multiple cameras using only one electronic compass, overcoming the drawback of existing technologies that require an electronic compass in each camera, leading to increased hardware costs. This method effectively reduces the hardware cost of multiple cameras by requiring only one electronic compass.

[0105] Furthermore, in related technologies, each camera in a multi-camera setup requires an electronic compass. The presence of multiple electronic compasses makes the multi-camera system more susceptible to external interference, resulting in lower accuracy and success rate of coordinate calibration. This invention calibrates the vertical coordinates of multiple cameras using only a single electronic compass, effectively avoiding the aforementioned interference issues caused by multiple electronic compasses, thus improving the accuracy and success rate of multi-camera coordinate calibration.

[0106] The multi-camera vertical coordinate calibration device provided by the present invention is described below. The multi-camera vertical coordinate calibration device described below and the multi-camera vertical coordinate calibration method described above can be referred to in correspondence.

[0107] Figure 13 This is a schematic diagram of the structure of the multi-camera vertical coordinate calibration device provided by the present invention. The multi-camera system includes a first camera and at least one second camera, and the first camera is equipped with an electronic compass.

[0108] like Figure 13 As shown, the multi-camera vertical coordinate calibration device 1300 includes: a control module 1301, an acquisition module 1302, a processing module 1303, and a calibration module 1304.

[0109] The control module 1301 is configured to control the first camera to face a preset direction and collect a first image and control the second camera to rotate from an initial position along a set direction and collect a second image in a vertical plane; and the lenses of the first camera and the second camera face the same direction in a horizontal direction.

[0110] The acquisition module 1302 is configured to acquire a rotation angle of the second camera and acquire a number of feature points in an overlapping area of the second image and the first image at each rotation angle.

[0111] The processing module 1303 is configured to obtain an overlapping feature point change curve according to a corresponding relationship between the rotation angle and the number of feature points in the overlapping area at each rotation angle.

[0112] The calibration module 1304 is configured to calibrate a vertical direction coordinate of the second camera according to the overlapping feature point change curve and a vertical direction coordinate of the first camera determined based on the electronic compass.

[0113] The multi-camera vertical direction coordinate calibration device provided by the application includes a first camera and at least one second camera, and the first camera is provided with an electronic compass. The control module 1301 is configured to control the first camera to face a preset direction and collect a first image and control the second camera to rotate from an initial position along a set direction and collect a second image in a vertical plane; and the lenses of the first camera and the second camera face the same direction in a horizontal direction. The acquisition module 1302 is configured to acquire a rotation angle of the second camera and acquire a number of feature points in an overlapping area of the second image and the first image at each rotation angle. The processing module 1303 is configured to obtain an overlapping feature point change curve according to a corresponding relationship between the rotation angle and the number of feature points in the overlapping area at each rotation angle. The calibration module 1304 is configured to calibrate a vertical direction coordinate of the second camera according to the overlapping feature point change curve and a vertical direction coordinate of the first camera determined based on the electronic compass. Therefore, the application can calibrate the vertical direction coordinates of the multi-camera with only one electronic compass in the multi-camera, solve the defects that the multi-camera needs to install an electronic compass in each camera in the prior art, increase the hardware cost of the multi-camera and be easily disturbed by external factors, retain only one electronic compass in the multi-camera, reduce the hardware cost of the multi-camera, effectively eliminate the interference of external factors, and improve the accuracy of the coordinate calibration result.

[0114] Optionally, the lenses of the first camera and the second camera have different field angles of view, and the acquisition module 1302 includes a first determination sub-module (not shown in the figure) configured to determine the number of feature points in the overlapping area of the second image and the first image at each rotation angle according to the rotation angle of the second camera. Figure 13the first image according to the field of view angles of the lenses of the first camera and the second camera; wherein the large field of view angle image corresponds to a camera with a larger field of view angle than the small field of view angle image; the second determining submodule (not shown) determines an effective matching region in the large field of view angle image according to the small field of view angle image; the processing submodule (not shown) obtains the number of feature points in the overlapping region between the effective matching region in the large field of view angle image and the small field of view angle image. Figure 13 Figure 13

[0115] Optionally, the second determining submodule is configured to: in the large field of view angle image, determine a region with the center of the large field of view angle image as the center and the imaging circle radius of the small field of view angle image as the radius as the effective matching region.

[0116] Optionally, the calibration module 1304 comprises: an angle determining submodule (not shown) configured to: according to the number of feature points in the overlapping region at each point on the overlapping feature point change curve, determine whether there is a first rotation angle in a period of the overlapping feature point change curve; wherein the number of feature points in the overlapping region corresponding to the first rotation angle is greater than the number of feature points in the overlapping region corresponding to other rotation angles in the period of the first rotation angle; a control submodule (not shown) configured to: in the case that there is the first rotation angle, rotate the second camera from the initial position by the first rotation angle in the set direction; a first adjusting submodule (not shown) configured to: after the control submodule rotates the second camera from the initial position by the first rotation angle in the set direction, adjust the vertical direction coordinate of the second camera to be consistent with the first camera according to the coordinate of the first camera in the vertical direction. Figure 13 Figure 13 Figure 13

[0117] Optionally, the calibration module 1304 further comprises: an inflection point determining submodule (not shown) configured to: in the case that there is no first rotation angle, determine a plurality of target inflection points in a period of the overlapping feature point change curve; wherein the plurality of target inflection points comprises: a first target inflection point when the number of feature points in the overlapping region on the overlapping feature point change curve increases and is greater than a set value as the rotation angle increases, and a second target inflection point when the number of feature points in the overlapping region on the overlapping feature point change curve decreases and is less than a set value as the rotation angle decreases; a calculation submodule (not shown) configured to: according to the reference rotation angle corresponding to each target inflection point in the plurality of target inflection points on the overlapping feature point change curve, calculate a second rotation angle; a rotating submodule (not shown) configured to: in the case that there is no first rotation angle, rotate the second camera from the initial position by the second rotation angle in the set direction. Figure 13 Figure 13 Figure 13 ​​​​​​​(Not shown in the image), used to rotate the second camera from its initial position along a set direction by a second rotation angle; second adjustment submodule ( Figure 13 (Not shown in the image) is used to adjust the vertical coordinates of the second camera to be consistent with those of the first camera, based on the vertical coordinates of the first camera, after the rotation submodule rotates the second camera from its initial position by a second rotation angle along a set direction.

[0118] Optionally, the computation submodule ( Figure 14 (Not shown in the text) is used to: sum the reference rotation angles corresponding to each target inflection point on the curve of the change of the coincident feature point among multiple target inflection points, and obtain the summation result; obtain the ratio of the summation result to the number of multiple target inflection points, and use the ratio as the second rotation angle.

[0119] Figure 14 An example is a schematic diagram of the physical structure of an electronic device, such as... ​ As shown, the electronic device may include: a processor 1410, a communications interface 1420, a memory 1430, and a communication bus 1440, wherein the processor 1410, the communications interface 1420, and the memory 1430 communicate with each other through the communication bus 1440. The processor 1410 can call logic instructions in the memory 1430 to execute a multi-camera vertical coordinate calibration method. The multi-camera includes a first camera and at least one second camera. The first camera is equipped with an electronic compass. The method includes: in a vertical plane, controlling the first camera to face a preset direction and acquire a first image, and controlling the second camera to rotate from an initial position along a set direction and acquire a second image; wherein the lenses of the first camera and the second camera face the same direction in the horizontal direction; acquiring the rotation angle of the second camera and acquiring the number of feature points in the overlapping area of ​​the second image and the first image at each rotation angle; obtaining a curve of overlapping feature point variation based on the correspondence between the rotation angle and the number of feature points in the overlapping area at each rotation angle; and calibrating the vertical coordinates of the second camera based on the curve of overlapping feature point variation and the vertical coordinates of the first camera determined by the electronic compass.

[0120] In addition, the logic instructions in the memory 1430 described above can be implemented in the form of software functional units and sold or used as independent products, and can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application essentially or the part that contributes to the prior art or part of the technical solutions can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a plurality of instructions to make a computer device (which can be a personal computer, a server, or a network device, etc.) execute all or part of the steps of the methods described in various embodiments of the present application. The foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various media that can store program codes.

[0121] In another aspect, the present application also provides a computer program product, the computer program product comprising a computer program, the computer program being stored in a computer readable storage medium, and the computer program being executed by a processor, so that the computer can execute the multi-camera vertical direction coordinate calibration method provided by the above-mentioned method. The multi-camera includes a first camera and at least one second camera, and the first camera is provided with an electronic compass. The method comprises: in a vertical plane, controlling the first camera to face a preset direction and collecting a first image, and controlling the second camera to rotate from an initial position along a set direction and collecting a second image; wherein the lenses of the first camera and the second camera are consistent in the horizontal direction; obtaining the rotation angle of the second camera, and obtaining the number of feature points in the overlapping area of the second image and the first image under each rotation angle; obtaining the overlapping feature point change curve according to the corresponding relationship between the rotation angle and the number of feature points in the overlapping area under each rotation angle; and calibrating the vertical direction coordinate of the second camera according to the overlapping feature point change curve and the vertical direction coordinate of the first camera determined based on the electronic compass.

[0122] In yet another aspect, the present application also provides a computer readable storage medium having stored thereon a computer program, which, when executed by a processor, implements a multi-camera vertical direction coordinate calibration method provided by any of the above methods, the multi-camera comprising a first camera and at least one second camera, the first camera being provided with an electronic compass, the method comprising: in a vertical plane, controlling the first camera to face a preset direction and capture a first image, and controlling the second camera to rotate from an initial position along a set direction and capture a second image; wherein the lenses of the first camera and the second camera are consistent in the horizontal direction; obtaining a rotation angle of the second camera, and obtaining the number of feature points in the overlapping area of the second image and the first image at each rotation angle; obtaining an overlapping feature point change curve according to the corresponding relationship between the rotation angle and the number of feature points in the overlapping area at each rotation angle; and calibrating the vertical direction coordinate of the second camera according to the overlapping feature point change curve and the vertical direction coordinate of the first camera determined based on the electronic compass.

[0123] The device embodiments described above are merely illustrative, wherein the units described as separate components can or can not be physically separate, and the components displayed as units can or can not be physical units, i.e., can be located in one place or distributed on multiple network units. Part or all of the modules can be selected to achieve the purpose of the embodiment scheme according to actual needs. Those skilled in the art can understand and implement without creative labor.

[0124] From the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be realized by means of software and the necessary general hardware platform, and of course can also be realized by hardware. Based on such understanding, the above technical solutions can be embodied in the form of a software product, which can be stored in a computer readable storage medium, such as a ROM / RAM, a magnetic disk, an optical disk, etc., and includes a plurality of instructions to make a computer device (which can be a personal computer, a server, or a network device, etc.) execute the methods described in each embodiment or some parts of the embodiments.

[0125] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement to some technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A multi-camera vertical direction coordinate calibration method, characterized by, The multi-camera includes a first camera and at least one second camera, and an electronic compass is arranged in the first camera, and the method comprises: In a vertical plane, the first camera is controlled to face a preset direction and capture a first image, and the second camera is controlled to rotate from an initial position along a set direction and capture a second image; wherein the lenses of the first camera and the second camera face in the same direction in a horizontal direction; An angle of rotation of the second camera is obtained, and the number of feature points in an overlapping area of the second image and the first image at each angle of rotation is obtained; According to a corresponding relationship between the angle of rotation and the number of feature points in the overlapping area at each angle of rotation, an overlapping feature point change curve is obtained; According to the overlapping feature point change curve and a vertical direction coordinate of the first camera determined based on the electronic compass, a vertical direction coordinate of the second camera is calibrated.

2. The multi-camera vertical direction coordinate calibration method of claim 1, wherein, The lenses of the first camera and the second camera have different field angles, and the number of feature points in the overlapping area of the second image and the first image at each angle of rotation comprises: According to the field angles of the lenses of the first camera and the second camera, a large field angle image and a small field angle image are determined in the second image and the first image; wherein the field angle of the camera corresponding to the large field angle image is greater than the field angle of the camera corresponding to the small field angle image; According to the small field angle image, an effective comparison area is determined in the large field angle image; The number of feature points in the overlapping area of the effective comparison area in the large field angle image and the small field angle image at each angle of rotation is obtained.

3. The multi-camera vertical direction coordinate calibration method of claim 2, wherein, The determination of the effective comparison area in the large field angle image according to the small field angle image comprises: In the large field angle image, an area with the center of the large field angle image as the center and the imaging circle radius of the small field angle image as the radius is determined as the effective comparison area.

4. The multi-camera vertical direction coordinate calibration method according to any one of claims 1-3, wherein, The calibration of the vertical direction coordinate of the second camera according to the overlapping feature point change curve and the vertical direction coordinate of the first camera determined based on the electronic compass comprises: According to the number of feature points in the overlapping area at each point on the overlapping feature point change curve, it is determined whether there is a first rotation angle in a period of the overlapping feature point change curve; wherein the number of feature points in the overlapping area corresponding to the first rotation angle is greater than the number of feature points in the overlapping area corresponding to other rotation angles in the period in which the first rotation angle is located; In the case where the first rotation angle exists, the second camera is rotated from the initial position along the set direction by the first rotation angle; According to the coordinate of the first camera in the vertical direction, the vertical direction coordinate of the second camera is adjusted to be consistent with that of the first camera.

5. The multi-camera vertical direction coordinate calibration method of claim 4, wherein, The calibration of the vertical direction coordinate of the second camera according to the overlapping feature point change curve and the vertical direction coordinate of the first camera determined based on the electronic compass further comprises: determining a plurality of target inflection points in one period of the coincident feature point change curve in the absence of the first rotation angle; wherein the plurality of target inflection points comprises a first target inflection point at which the number of feature points in the coincident region on the coincident feature point change curve increases and is greater than a set value as the rotation angle increases, and a second target inflection point at which the number of feature points in the coincident region on the coincident feature point change curve decreases and is less than the set value as the rotation angle decreases; calculating a second rotation angle according to a reference rotation angle corresponding to each of the plurality of target inflection points on the coincident feature point change curve; rotating the second camera from the initial position by the second rotation angle in the set direction; adjusting the vertical direction coordinate of the second camera to be consistent with that of the first camera according to the vertical direction coordinate of the first camera.

6. The multi-camera vertical direction coordinate calibration method of claim 5, wherein, The calculating a second rotation angle according to a reference rotation angle corresponding to each of the plurality of target inflection points on the coincident feature point change curve comprises: summing the reference rotation angles corresponding to each of the plurality of target inflection points on the coincident feature point change curve to obtain a summation result; obtaining a ratio of the summation result to the number of the plurality of target inflection points, and taking the ratio as the second rotation angle.

7. A multi-camera vertical coordinate calibration device, characterized in that, The multi-camera comprises a first camera and at least one second camera, and the first camera is provided with an electronic compass, and the device comprises: a control module configured to control the first camera to face a preset direction and capture a first image, and control the second camera to rotate from an initial position in a set direction and capture a second image in a vertical plane, wherein the lenses of the first camera and the second camera face in the same direction in a horizontal direction; an acquisition module configured to acquire rotation angles of the second camera, and acquire the number of feature points in a coincident region of the second image and the first image at each rotation angle; a processing module configured to obtain a coincident feature point change curve according to a corresponding relationship between the rotation angles and the number of feature points in the coincident region at each rotation angle; a calibration module configured to calibrate a vertical direction coordinate of the second camera according to the coincident feature point change curve and the vertical direction coordinate of the first camera determined based on the electronic compass.

8. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor implements the multi-camera vertical direction coordinate calibration method of any one of claims 1 to 6 when executing the program.

9. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program implements the multi-camera vertical direction coordinate calibration method of any one of claims 1 to 6 when executed by the processor.

10. A computer program product comprising a computer program, characterized in that, The computer program implements the multi-camera vertical direction coordinate calibration method of any one of claims 1 to 6 when executed by the processor.

Images