Method, device and electronic equipment for detecting camera data quality

Abstract

The application relates to the technical field of cameras, in particular to a camera data quality detection method and device and electronic equipment, the method comprising the following steps: acquiring initial point clouds collected by a target camera from a target region, the target region comprising at least one plane reference object; performing clustering processing based on the initial point clouds to obtain target point clouds corresponding to each plane reference object; fitting each target point cloud to obtain a fitting result; determining a data quality detection result of the target camera based on the relationship between the attributes of each fitting result and a preset condition and the distance between the fitting result and the target camera, wherein the preset condition is determined based on the relationship between the plane reference object and the target camera. The fitting result and the actual arrangement of the plane reference object are used for data detection, wherein the attributes of the fitting result are used for detecting data distortion, and the distance is used for detecting distance error, so that the detection efficiency is improved.

Description

Technical Field

[0001] This invention relates to the field of camera technology, and more specifically to methods, apparatus, and electronic devices for detecting camera data quality. Background Technology

[0002] When manufacturing equipment that uses sensors for 3D spatial perception, such as depth cameras, it is necessary to test the quality of the camera data measured by these devices to avoid problems with camera data quality caused by factory quality issues or factory installation, such as camera deformation due to fixing structures or unqualified lens installation, which could affect the use of subsequent products.

[0003] Existing detection methods generally rely on manual inspection. For example, they acquire point cloud data of a fixed object and determine the quality of the camera data by checking whether the distances and shapes of the point cloud data meet expectations. However, this method depends on subjective human judgment, lacks consistency, and is inefficient. Summary of the Invention

[0004] In view of this, embodiments of the present invention provide a method, apparatus and electronic device for detecting camera data quality, so as to solve the problems of inconsistent and inefficient camera data quality detection methods.

[0005] According to a first aspect, embodiments of the present invention provide a method for detecting camera data quality, comprising:

[0006] Acquire an initial point cloud of a target region obtained by a target camera, wherein the target region includes at least one planar reference object;

[0007] Clustering is performed on the initial point cloud to obtain target point clouds that correspond one-to-one with each of the planar reference objects;

[0008] The fitting results are obtained by fitting each of the target point clouds;

[0009] Based on the attributes of each fitting result and the relationship between the distance between the fitting result and the target camera and the corresponding preset conditions, the data quality detection result of the target camera is determined, wherein the preset conditions are determined based on the relationship between the planar reference object and the target camera.

[0010] The camera data quality detection method provided in this embodiment of the invention clusters the initial point cloud to obtain target point clouds corresponding to each planar reference object, then fits the target point cloud to obtain a fitting result, and uses the fitting result and the actual arrangement of the planar reference objects to detect the data. The attributes of the fitting result are used to detect data distortion, and the distance is used to detect distance error. Based on these two aspects, reliable detection of camera data quality can be achieved, improving detection efficiency.

[0011] In conjunction with the first aspect, in the first embodiment of the first aspect, the step of performing clustering processing based on the initial point cloud to obtain a target point cloud corresponding one-to-one with each of the planar reference objects includes:

[0012] Obtain the height range of the planar reference object in the camera coordinate system of the target camera;

[0013] The initial point cloud is filtered based on the height range to obtain the first point cloud;

[0014] Clustering is performed on the first point cloud to obtain target point clouds that correspond one-to-one with each of the planar reference objects.

[0015] The camera data quality detection method provided in this embodiment of the invention uses the height range of a planar reference object in the camera coordinate system to filter the initial point cloud, so that all point clouds of the planar reference object are in the target point cloud and irrelevant point clouds are excluded. On the one hand, this reduces the amount of data processing, and on the other hand, it ensures the reliability of the obtained target point cloud.

[0016] In conjunction with the first embodiment of the first aspect, in the second embodiment of the first aspect, the step of performing clustering processing based on the first point cloud to obtain a target point cloud corresponding one-to-one with each of the planar reference objects includes:

[0017] The first point cloud is projected onto the plane containing the optical axis of the target camera to obtain the second point cloud;

[0018] Clustering is performed on the second point cloud to obtain target point clouds that correspond one-to-one with each of the aforementioned planar reference objects.

[0019] The camera data quality detection method provided in this embodiment of the invention projects the first point cloud to map the three-dimensional first point cloud into a two-dimensional second point cloud, so as to facilitate subsequent line fitting and ensure the reliability of line fitting.

[0020] In conjunction with the second embodiment of the first aspect, in the third embodiment of the first aspect, the step of performing clustering processing based on the second point cloud to obtain a target point cloud corresponding one-to-one with each of the planar reference objects includes:

[0021] For each current data point in the second point cloud, count the number of other data points in the first point cloud whose distance to the current data point is less than a preset distance;

[0022] When the number is less than a preset number, the current data point is removed from the second point cloud to obtain the third point cloud;

[0023] Clustering is performed on the third point cloud to obtain target point clouds that correspond one-to-one with each of the planar reference objects.

[0024] The camera data quality detection method provided in this embodiment of the invention can remove outliers to obtain an accurate third point cloud by removing discrete points from the second point cloud, thus ensuring the accuracy of the obtained target point cloud.

[0025] In conjunction with the first aspect, or any one of the first to third embodiments of the first aspect, in the fourth embodiment of the first aspect, when the target point cloud is a point cloud on the plane where the optical axis of the target camera is located, the step of fitting each of the target point clouds to obtain a fitting result includes:

[0026] A straight line is fitted to each of the target point clouds to obtain the fitted straight line corresponding to each target point cloud.

[0027] The camera data quality detection method provided in this embodiment of the invention performs linear fitting on the target point cloud when mapped to the plane where the optical axis is located, which simplifies data processing and improves detection efficiency.

[0028] In conjunction with the fourth embodiment of the first aspect, in the fifth embodiment of the first aspect, determining the data quality detection result of the target camera based on the attributes of each of the fitting results and the relationship between the distance between the fitting result and the target camera and the corresponding preset conditions includes:

[0029] Obtain the slope of each of the fitted lines;

[0030] Calculate the first distance between each of the fitted lines and the target camera;

[0031] When both the slope and the first distance meet the corresponding preset conditions, the data quality of the target camera is determined to be qualified.

[0032] The camera data quality detection method provided in this embodiment of the invention indicates that the point cloud data distortion meets the requirements and the ranging error meets the requirements when both the slope and the first distance meet the corresponding preset conditions, thus detecting the camera data quality from two aspects.

[0033] In conjunction with the fifth embodiment of the first aspect, in the sixth embodiment of the first aspect, determining the data quality detection result of the target camera based on the attributes of each of the fitting results and the relationship between the distance between the fitting result and the target camera and the corresponding preset conditions further includes:

[0034] Obtain the target point cloud corresponding to each of the fitted lines;

[0035] For each target data point in the target point cloud, calculate the second distance between the target data point and the corresponding fitted line;

[0036] When the slope, the first distance, and the second distance all meet the corresponding preset conditions, the data quality of the target camera is determined to be qualified.

[0037] The camera data quality detection method provided in this embodiment of the invention judges the data quality by using a second distance. When the second distance meets the corresponding preset conditions, it indicates that the camera distance noise meets the requirements, thereby further detecting the quality of the camera data and improving the reliability of the detection results.

[0038] According to a second aspect, embodiments of the present invention also provide a camera data quality detection device, comprising:

[0039] The acquisition module is used to acquire the initial point cloud obtained by the target camera from the target area, wherein the target area includes at least one planar reference object;

[0040] The clustering module is used to perform clustering processing based on the initial point cloud to obtain target point clouds that correspond one-to-one with each of the planar reference objects;

[0041] The fitting module is used to fit each of the target point clouds to obtain the fitting result;

[0042] A determination module is used to determine the data quality detection result of the target camera based on the attributes of each of the fitting results and the relationship between the distance between the fitting result and the target camera and the corresponding preset conditions. The preset conditions are determined based on the relationship between the planar reference object and the target camera.

[0043] According to a third aspect, embodiments of the present invention provide an electronic device, including: a memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the computer instructions to perform the camera data quality detection method described in the first aspect or any embodiment of the first aspect.

[0044] According to a fourth aspect, embodiments of the present invention provide a computer-readable storage medium storing computer instructions for causing the computer to perform the camera data quality detection method described in the first aspect or any embodiment of the first aspect.

[0045] It should be noted that the corresponding beneficial effects of the camera data quality detection device, electronic device and computer-readable storage medium provided in the embodiments of the present invention can be found in the description of the corresponding beneficial effects in the camera data quality detection method above, and will not be repeated here. Attached Figure Description

[0046] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0047] Figures 1a-1b This is a schematic diagram of a camera data quality detection system according to an embodiment of the present invention;

[0048] Figure 2 This is a flowchart of a camera data quality detection method according to an embodiment of the present invention;

[0049] Figure 3 This is a flowchart of a camera data quality detection method according to an embodiment of the present invention;

[0050] Figure 4 This is a schematic diagram of a camera data quality detection system according to an embodiment of the present invention;

[0051] Figure 5 This is a flowchart of a camera data quality detection method according to an embodiment of the present invention;

[0052] Figure 6 This is a structural block diagram of a camera data quality detection device according to an embodiment of the present invention;

[0053] Figure 7 This is a schematic diagram of the hardware structure of the electronic device provided in an embodiment of the present invention. Detailed Implementation

[0054] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0055] An optional detection scenario for the camera data quality detection method provided in this embodiment of the invention is as follows: Figure 1a as well as Figure 1b As shown, where, Figure 1a This is a side view of the detection scene. Figure 1b This is a top view of the detection scene. The scene includes a target camera 10 and at least one planar reference object 20. The planar reference object 20 can be any planar object, and its form is not limited here. When arranging each planar reference object 20, it can be perpendicular to the camera's optical axis (i.e., the Z-axis) or at a certain angle to the camera's optical axis; there are no restrictions on this. Simultaneously, the planar reference objects 20 are arranged to ensure that they can all be captured by the target camera without obstructing each other. After the positions of each planar reference object 20 are fixed, the distance between each planar reference object 20 and the target camera can be determined using a distance-based device.

[0056] It should be noted that the various planar reference objects 20 can be the same or different, etc. To reduce the amount of data processing during detection, the same planar reference object 20 can be used; or, to improve the detection accuracy, different planar reference objects 20 can be used. The selection of each planar reference object 20 is set according to the actual detection scenario requirements.

[0057] According to an embodiment of the present invention, a method for detecting camera data quality is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0058] This embodiment provides a method for detecting camera data quality, which can be used in electronic devices such as computers, tablets, and servers. Figure 2 This is a flowchart of a camera data quality detection method according to an embodiment of the present invention, such as... Figure 2 As shown, the process includes the following steps:

[0059] S11, acquire the initial point cloud obtained by the target camera from the target area.

[0060] The target area includes at least one planar reference object.

[0061] The target area is an area with at least one planar reference object. For details on the arrangement and types of planar reference objects, please refer to the above description; they will not be repeated here. After acquiring an initial point cloud, the target camera sends the initial point cloud to the electronic device, which then acquires the initial point cloud. Alternatively, the target camera acquires multiple raw point clouds of the target area and sends them to the electronic device. The electronic device performs image quality detection on the multiple raw point clouds and selects the raw point cloud with the highest image quality as the initial point cloud.

[0062] S12, perform clustering based on the initial point cloud to obtain target point clouds that correspond one-to-one with each planar reference object.

[0063] Since the initial point cloud is a point cloud of at least one planar reference object, it is necessary to determine the target point clouds belonging to each planar reference object from the initial point cloud. The electronic device can determine each target point cloud by performing clustering processing on the initial point clouds; alternatively, it can remove outliers from the initial point clouds before performing clustering. Specifically, the electronic device can use the height range of the planar reference objects to filter the initial point clouds, or it can use an outlier filter to remove outliers from the initial point clouds, and so on.

[0064] The specifics of this step will be described in detail below.

[0065] S13, Fit the point cloud of each target to obtain the fitting result.

[0066] The electronic device can perform either linear fitting or planar fitting. Specifically, when linear fitting is chosen, if the planar reference is perpendicular to the optical axis of the target camera, the target point cloud can be directly projected onto the plane containing the optical axis, for example, the X=0 plane, resulting in the Y=0 plane. Linear fitting is then performed on the projection to obtain the fitting result. If there is an angle between the planar reference and the optical axis of the target camera, relative plane fitting is required. The target point cloud is rotated and reprojected using the relative rotation relationship between the plane's normal vector and the Z=0 plane normal (0, 0, 1), and then linear fitting is performed on the projection to obtain the fitting result. When planar fitting is chosen, the electronic device directly performs planar fitting on each target point cloud to obtain the fitting plane.

[0067] Whether to choose linear fitting or planar fitting is determined by actual needs, and no restrictions are imposed on this alignment.

[0068] S14. Based on the attributes of each fitting result and the relationship between the distance between the fitting result and the target camera and the corresponding preset conditions, determine the data quality detection result of the target camera.

[0069] The preset conditions are determined based on the relationship between the planar reference object and the target camera.

[0070] The attributes of the fitting result for linear fitting include the slope of the fitted line; the attributes of the fitting result for planar fitting include the normal vector of the fitted plane. Both linear and planar fitting are performed by calculating the distance from the origin of the camera coordinate system to the fitted line or the fitted plane.

[0071] After obtaining the attributes and distances, the electronic device compares these attributes and distances with corresponding preset conditions. Specifically, if the data is normal, the fitted line should be perpendicular to the Z-axis, i.e., parallel to the X-axis. Correspondingly, the slope of the fitted line should be less than a preset value; the distance between the origin and the fitted line should be within a certain range of the distance between the origin and the planar reference object measured in S11. Similarly, the normal vector and distance of the fitted plane also need to be compared separately.

[0072] When electronic equipment undergoes quality inspection, the camera measurement is compared with the actual parameters of the plane reference object to assess whether there are distance errors and distortions, i.e. whether they exceed the allowable range, thereby obtaining the inspection result.

[0073] The camera data quality detection method provided in this embodiment clusters the initial point cloud to obtain target point clouds corresponding to each planar reference object. Then, it fits the target point cloud to obtain the fitting result. The data is detected by comparing the fitting result with the actual arrangement of the planar reference objects. The attributes of the fitting result are used to detect data distortion, and the distance is used to detect distance error. Based on these two aspects, reliable detection of camera data quality can be achieved, thus improving detection efficiency.

[0074] This embodiment provides a method for detecting camera data quality, which can be used in electronic devices such as computers, tablets, and servers. Figure 3 This is a flowchart of a camera data quality detection method according to an embodiment of the present invention, such as... Figure 3 As shown, the process includes the following steps:

[0075] S21, acquire the initial point cloud obtained by the target camera from the target area.

[0076] The target area includes at least one planar reference object.

[0077] Please see details Figure 2 S11 of the illustrated embodiment will not be described again here.

[0078] S22, clustering is performed based on the initial point cloud to obtain the target point cloud that corresponds one-to-one with each planar reference object.

[0079] Specifically, S22 includes:

[0080] S221, obtain the height range of the planar reference object in the camera coordinate system of the target camera.

[0081] Since the target camera's acquisition range is known, and the heights of various planar reference objects are also obtainable, the height range of the planar reference objects in the camera coordinate system can be obtained by mapping them from the world coordinate system to the camera coordinate system. For example, ... Figure 4 As shown, the heights of the highest and lowest points of the planar reference object in the camera coordinate system are obtained, thus yielding the corresponding height range.

[0082] The electronic device filters out the highest and lowest points from the height range corresponding to all planar references, and determines the distance between the highest and lowest points as the height range in the camera coordinate system.

[0083] S222, the initial point cloud is filtered based on the height range to obtain the first point cloud.

[0084] Since the initial point cloud includes point clouds of other objects besides the planar reference object, and this method analyzes the planar reference object, it is necessary to exclude the point clouds of other objects as much as possible. Therefore, for the height range determined in S221 above, the point cloud height of the planar reference object is within this range. Thus, the electronic device uses this height range to filter the initial point cloud, removing point clouds outside the height range to obtain the first point cloud.

[0085] S223, perform clustering based on the first point cloud to obtain target point clouds that correspond one-to-one with each planar reference object.

[0086] The electronic device uses a clustering algorithm to calculate the distance to the first point cloud, obtaining the target point cloud belonging to each planar reference object. Of course, before clustering, the electronic device can also remove discrete points from the first point cloud. Alternatively, if straight line fitting is to be performed, the first point cloud needs to be mapped to the plane containing the optical axis.

[0087] In some alternative implementations, S223 includes:

[0088] (1) Project the first point cloud onto the plane of the optical axis of the target camera to obtain the second point cloud.

[0089] (2) Clustering is performed based on the second point cloud to obtain the target point cloud that corresponds one-to-one with each planar reference object.

[0090] The plane containing the optical axis can be either the X=0 plane or the Y=0 plane in the camera coordinate system. Therefore, if projected onto the X=0 plane, the electronic device sets the x-value of each data point in the first point cloud to 0; if projected onto the Y=0 plane, the electronic device sets the y-value of each data point in the first point cloud to 0. After projection processing, the second point cloud is obtained.

[0091] The electronic device performs clustering processing on the second point cloud to obtain the target point cloud on a two-dimensional plane. The first point cloud is then projected to map the three-dimensional first point cloud into the two-dimensional second point cloud, which facilitates subsequent line fitting and ensures the reliability of the line fitting.

[0092] In some alternative implementations, step (2) of S223 above includes:

[0093] 2.1) For each current data point in the second point cloud, count the number of other data points in the first point cloud whose distance to the current data point is less than a preset distance.

[0094] 2.2) When the number is less than the preset number, remove the current data point from the second point cloud to obtain the third point cloud.

[0095] The electronic device uses an outlier filter to remove outliers from the second point cloud, resulting in a third point cloud. Specifically, for each point in the second point cloud, it checks whether there are n or more points in the second point cloud whose distances to that point are less than r. If so, the point is retained; otherwise, it is removed. Here, n and r need to be determined based on the target camera resolution, the plane height of the planar reference object, and the voxel filter resolution. In voxel filtering, the concept of a voxel is similar to a pixel; a pixel is a two-dimensional point, while a voxel is a small three-dimensional space.

[0096] 2.3) Cluster the third point cloud to obtain the target point cloud that corresponds one-to-one with each planar reference object.

[0097] After removing discrete points from the second point cloud, the electronic device obtains a third point cloud. This third point cloud is then clustered to obtain the target point cloud. By removing discrete points from the second point cloud, outliers can be eliminated to obtain an accurate third point cloud, ensuring the accuracy of the resulting target point cloud.

[0098] S23, Fit the point cloud of each target to obtain the fitting result.

[0099] Please see details Figure 2 S13 of the illustrated embodiment will not be described again here.

[0100] S24. Based on the attributes of each fitting result and the relationship between the distance between the fitting result and the target camera and the corresponding preset conditions, determine the data quality detection result of the target camera.

[0101] The preset conditions are determined based on the relationship between the planar reference object and the target camera.

[0102] Please see details Figure 2 S14 of the illustrated embodiment will not be described again here.

[0103] The camera data quality detection method provided in this embodiment uses the height range of the planar reference object in the camera coordinate system to filter the initial point cloud, so that all the point clouds of the planar reference object are in the target point cloud and irrelevant point clouds are excluded. On the one hand, it reduces the amount of data processing, and on the other hand, it ensures the reliability of the obtained target point cloud.

[0104] This embodiment provides a method for detecting camera data quality, which can be used in electronic devices such as computers, tablets, and servers. Figure 5 This is a flowchart of a camera data quality detection method according to an embodiment of the present invention, such as... Figure 5 As shown, the process includes the following steps:

[0105] S31, acquire the initial point cloud obtained by the target camera from the target area.

[0106] The target area includes at least one planar reference object.

[0107] Please see details Figure 2 S11 of the illustrated embodiment will not be described again here.

[0108] S32, clustering is performed based on the initial point cloud to obtain target point clouds that correspond one-to-one with each planar reference object.

[0109] Please see details Figure 3 S22 of the illustrated embodiment will not be described again here.

[0110] S33, fit the point cloud of each target to obtain the fitting result.

[0111] In this embodiment, the electronic device obtains the fitted line corresponding to each target point cloud by performing linear fitting on each target point cloud. Specifically, linear fitting algorithms, such as RANSAC and least squares methods, are used to perform linear fitting to obtain the linear equation of the line corresponding to each target point cloud: y = kx + b. The attributes of the fitting result include: {{k1,b1},{k2,b2},{k3,b3},...}.

[0112] S34. Based on the attributes of each fitting result and the relationship between the distance between the fitting result and the target camera and the corresponding preset conditions, determine the data quality detection result of the target camera.

[0113] The preset conditions are determined based on the relationship between the planar reference object and the target camera.

[0114] Specifically, S34 includes:

[0115] S341, obtain the slope of each fitted line.

[0116] As described above, the slopes of the respective fitting lines are k1, k2, k3, ….

[0117] S342. Calculate the first distance between each fitting line and the target camera.

[0118] For each fitting line, the electronic device calculates the distance between the origin of the camera coordinate system of the target camera and the fitting line to obtain the first distance.

[0119] S343. When both the slope and the first distance meet the corresponding preset conditions, determine that the data quality of the target camera is qualified.

[0120] Since during line fitting, the point cloud is projected onto a plane perpendicular to the Z-axis in the camera coordinate system, if the data is normal, the fitted line should be perpendicular to the Z-axis, that is, parallel to the X-axis.

[0121] For each fitting line, the judgment qualification conditions are:

[0122] (a) |k| < k_thr; where, generally, global distortion causes bending at the edges. According to actual requirements, k_thr is the slope of the fitting line on the farthest plane reference object in the case of allowable maximum distortion at the edges.

[0123] (b) Given that the distances between the plane reference object and the origin of the camera coordinate system are (d1, d2, d3,...), find the value db that is closest to b of each fitting line among them, and |b - db| < db * e; where e is the percentage of the target camera accuracy error (i.e., performance parameter).

[0124] If the above two conditions are met, determine that the data quality of the target camera is qualified.

[0125] In some optional implementation manners, the above S34 further includes:

[0126] (1) Obtain the target point cloud corresponding to each fitting line.

[0127] (2) For each target data point in the target point cloud, calculate the second distance between the target data point and the corresponding fitting line.

[0128] (3) When the slope, the first distance, and the second distance all meet the corresponding preset conditions, determine that the data quality of the target camera is qualified.

[0129] For each target point cloud {P_c1, P_c2, P_c3,...} and its corresponding fitted line {{k1, b1}, {k2, b2}, {k3, b3},...}, calculate the reprojection error, that is, calculate the second distance from each point in P_ci to the corresponding fitted line y = ki + bi, and calculate the standard deviation based on this second distance. When the standard deviation < bi * e, the task is qualified. Or, variance calculation can also be performed based on the second distance, etc., and no limitations are imposed on this here.

[0130] Judgment is made through the second distance. When the second distance meets the corresponding preset condition, it indicates that the camera distance noise meets the requirements, thereby further detecting the quality of the camera data and further improving the reliability of the detection result.

[0131] The method for detecting the quality of camera data provided in this embodiment indicates that the distortion condition of the point cloud data meets the requirements and the ranging error meets the requirements when both the slope and the first distance meet the corresponding preset conditions, and the quality of the camera data is detected from two aspects.

[0132] In this embodiment, a device for detecting the quality of camera data is also provided. This device is used to implement the above embodiment and the preferred implementation manners, and those that have been described will not be repeated. As used below, the term "module" can be a combination of software and / or hardware that can achieve a predetermined function. Although the device described in the following embodiments is preferably implemented in software, implementation in hardware, or a combination of software and hardware is also possible and contemplated.

[0133] This embodiment provides a device for detecting the quality of camera data, as Figure 6 shown, including:

[0134] An acquisition module 41, configured to acquire the initial point cloud collected by the target camera for the target area, where the target area includes at least one planar reference object;

[0135] A clustering module 42, configured to perform clustering processing based on the initial point cloud to obtain target point clouds corresponding to each of the planar reference objects;

[0136] A fitting module 43, configured to fit each of the target point clouds to obtain a fitting result;

[0137] A determination module 44, configured to determine the detection result of the data quality of the target camera based on the relationship between the attributes of each of the fitting results and the distance between the fitting result and the target camera and the corresponding preset conditions, where the preset conditions are determined based on the relationship between the planar reference object and the target camera.

[0138] In some alternative implementations, clustering module 42 includes:

[0139] The first acquisition unit is used to acquire the height range of the planar reference object in the camera coordinate system of the target camera;

[0140] A filtering unit is used to filter the initial point cloud based on the height range to obtain a first point cloud;

[0141] Clustering unit, used to perform clustering processing based on the first point cloud to obtain target point clouds that correspond one-to-one with each of the planar reference objects.

[0142] In some alternative implementations, the clustering unit includes:

[0143] The projection subunit is used to project the first point cloud onto the plane where the optical axis of the target camera is located to obtain the second point cloud;

[0144] The clustering subunit is used to perform clustering processing based on the second point cloud to obtain target point clouds that correspond one-to-one with each of the planar reference objects.

[0145] In some alternative implementations, the clustering subunit includes:

[0146] The statistics subunit is used to count, for each current data point in the second point cloud, the number of other data points in the first point cloud whose distance to the current data point is less than a preset distance;

[0147] The sub-unit is used to remove the current data point from the second point cloud when the number is less than a preset number, so as to obtain the third point cloud;

[0148] The processing subunit is used to perform clustering processing on the third point cloud to obtain target point clouds that correspond one-to-one with each of the planar reference objects.

[0149] In some optional implementations, when the target point cloud is a point cloud on the plane where the optical axis of the target camera is located, the fitting module 43 includes:

[0150] The fitting unit is used to perform linear fitting on each of the target point clouds to obtain the fitted straight line corresponding to each of the target point clouds.

[0151] In some alternative implementations, the determining module 44 includes:

[0152] The second acquisition unit is used to acquire the slope of each of the fitted lines;

[0153] The first calculation unit is used to calculate the first distance between each of the fitted straight lines and the target camera;

[0154] The first determining unit is used to determine that the data quality of the target camera is qualified when both the slope and the first distance meet the corresponding preset conditions.

[0155] In some alternative implementations, the determining module 44 further includes:

[0156] The third acquisition unit is used to acquire the target point cloud corresponding to each of the fitted lines;

[0157] The second calculation unit is used to calculate the second distance between each target data point in the target point cloud and the corresponding fitted line;

[0158] The second determining unit is used to determine that the data quality of the target camera is qualified when the slope, the first distance, and the second distance all meet the corresponding preset conditions.

[0159] In this embodiment, the camera data quality detection device is presented in the form of a functional unit. Here, a unit refers to an ASIC circuit, a processor and memory that execute one or more software or fixed programs, and / or other devices that can provide the above functions.

[0160] Further functional descriptions of the above modules are the same as those in the corresponding embodiments described above, and will not be repeated here.

[0161] This invention also provides an electronic device having the above-described features. Figure 6 The device shown is for detecting camera data quality.

[0162] Please see Figure 7 , Figure 7 This is a schematic diagram of the structure of an electronic device provided in an optional embodiment of the present invention, such as... Figure 7 As shown, the electronic device may include: at least one processor 51, such as a CPU (Central Processing Unit), at least one communication interface 53, memory 54, and at least one communication bus 52. The communication bus 52 is used to enable communication between these components. The communication interface 53 may include a display screen or a keyboard; optionally, the communication interface 53 may also include a standard wired interface or a wireless interface. The memory 54 may be high-speed RAM (Random Access Memory) or non-volatile memory, such as at least one disk storage device. Optionally, the memory 54 may also be at least one storage device located remotely from the aforementioned processor 51. The processor 51 may be combined with... Figure 6The described apparatus has an application program stored in memory 54, and the processor 51 calls the program code stored in memory 54 to perform any of the above method steps.

[0163] The communication bus 52 can be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The communication bus 52 can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 7 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0164] The memory 54 may include volatile memory, such as random-access memory (RAM); the memory may also include non-volatile memory, such as flash memory, hard disk drive (HDD) or solid-state drive (SSD); the memory 54 may also include a combination of the above types of memory.

[0165] The processor 51 can be a central processing unit (CPU), a network processor (NP), or a combination of CPU and NP.

[0166] The processor 51 may further include a hardware chip. This hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The PLD may be a complex programmable logic device (CPLD), a field-programmable gate array (FPGA), a generic array logic (GAL), or any combination thereof.

[0167] Optionally, the memory 54 is also used to store program instructions. The processor 51 can invoke the program instructions to implement the camera data quality detection method as shown in any embodiment of this application.

[0168] This invention also provides a non-transitory computer storage medium storing computer-executable instructions that can execute the camera data quality detection method in any of the above method embodiments. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk drive (HDD), or solid-state drive (SSD), etc.; the storage medium may also include combinations of the above types of memory.

[0169] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A method for detecting camera data quality, characterized in that, include: Acquire an initial point cloud of a target region obtained by a target camera, wherein the target region includes at least one planar reference object; Clustering is performed on the initial point cloud to obtain target point clouds that correspond one-to-one with each of the planar reference objects; The fitting results are obtained by fitting each of the target point clouds; Based on the attributes of each fitting result and the relationship between the distance between the fitting result and the target camera and the corresponding preset conditions, the data quality detection result of the target camera is determined, wherein the preset conditions are determined based on the relationship between the planar reference object and the target camera; The step of determining the data quality detection result of the target camera based on the attributes of each fitting result and the relationship between the distance between the fitting result and the target camera and the corresponding preset conditions includes: The slope of each fitted line is obtained, wherein the fitted line is obtained by performing linear fitting on the target point cloud; Calculate the first distance between each of the fitted lines and the target camera; When both the slope and the first distance meet the corresponding preset conditions, the data quality of the target camera is determined to be qualified.

2. The method according to claim 1, characterized in that, The clustering process based on the initial point cloud to obtain target point clouds corresponding one-to-one with each of the planar reference objects includes: Obtain the height range of the planar reference object in the camera coordinate system of the target camera; The initial point cloud is filtered based on the height range to obtain the first point cloud; Clustering is performed on the first point cloud to obtain target point clouds that correspond one-to-one with each of the planar reference objects.

3. The method according to claim 2, characterized in that, The step of clustering based on the first point cloud to obtain target point clouds corresponding one-to-one with each of the planar reference objects includes: The first point cloud is projected onto the plane containing the optical axis of the target camera to obtain the second point cloud; Clustering is performed on the second point cloud to obtain target point clouds that correspond one-to-one with each of the aforementioned planar reference objects.

4. The method according to claim 3, characterized in that, The clustering process based on the second point cloud to obtain target point clouds corresponding one-to-one with each of the planar reference objects includes: For each current data point in the second point cloud, count the number of other data points in the second point cloud whose distance to the current data point is less than a preset distance; When the number is less than a preset number, the current data point is removed from the second point cloud to obtain the third point cloud; Clustering is performed on the third point cloud to obtain target point clouds that correspond one-to-one with each of the planar reference objects.

5. The method according to any one of claims 1-4, characterized in that, When the target point cloud is a point cloud on the plane where the optical axis of the target camera is located, the fitting of each of the target point clouds to obtain the fitting result includes: A straight line is fitted to each of the target point clouds to obtain the fitted straight line corresponding to each target point cloud.

6. The method according to claim 1, characterized in that, The method of determining the data quality detection result of the target camera based on the attributes of each fitting result and the relationship between the distance between the fitting result and the target camera and the corresponding preset conditions further includes: Obtain the target point cloud corresponding to each of the fitted lines; For each target data point in the target point cloud, calculate the second distance between the target data point and the corresponding fitted line; When the slope, the first distance, and the second distance all meet the corresponding preset conditions, the data quality of the target camera is determined to be qualified.

7. A camera data quality detection device, characterized in that, include: The acquisition module is used to acquire the initial point cloud obtained by the target camera from the target area, wherein the target area includes at least one planar reference object; The clustering module is used to perform clustering processing based on the initial point cloud to obtain target point clouds that correspond one-to-one with each of the planar reference objects; The fitting module is used to fit each of the target point clouds to obtain the fitting result; A determination module is used to determine the data quality detection result of the target camera based on the attributes of each of the fitting results and the relationship between the distance between the fitting result and the target camera and the corresponding preset conditions. The preset conditions are determined based on the relationship between the planar reference object and the target camera. The determining module includes: The third acquisition unit is used to acquire the slope of each fitted line, which is obtained by fitting the target point cloud with a straight line. The second calculation unit is used to calculate the first distance between each of the fitted lines and the target camera; The second determining unit is used to determine that the data quality of the target camera is qualified when both the slope and the first distance meet the corresponding preset conditions.

8. An electronic device, characterized in that, include: A memory and a processor are communicatively connected, the memory stores computer instructions, and the processor executes the computer instructions to perform the camera data quality detection method according to any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing the computer to perform the camera data quality detection method according to any one of claims 1-6.

Images