Data Processing Method, Apparatus, Computer Device, and Storage Medium

Abstract

This application relates to the fields of data processing and robotics, and discloses a data processing method, apparatus, computer device, and computer storage medium for effectively screening out noise data. The method part includes: obtaining a target data set, where the target data set includes a plurality of point data continuously detected by a radar; obtaining error information between all adjacent point data among the plurality of point data; grouping the plurality of point data according to the error information to obtain at least one data group; determining whether the point data in the data group is noise data according to the data information of the point data in the data group, and removing the point data determined to be noise data.

Description

Technical Field

[0001] This application relates to the technical field of data processing, and in particular to a data processing method, apparatus, computer device, and computer storage medium. Background Art

[0002] LiDAR is an important component in a robot. The robot obtains precise position information through the radar data detected by the LiDAR deployed on itself. The recognition accuracy of the robot for obstacles at different distances will directly affect the actual use effect. In the traditional solution, the radar data obtained by the robot is not effectively filtered for noise, and the radar data is easily interfered by noise points, resulting in low data accuracy. Summary of the Invention

[0003] Embodiments of this application provide a data processing method, apparatus, computer device, and storage medium to solve the technical problem that radar data in the traditional solution is easily interfered by noise points and has low data accuracy.

[0004] The first aspect of this application provides a data processing method, and the method includes:

[0005] Obtain a target data set, where the target data set includes a plurality of point data continuously detected by a radar;

[0006] Obtain the error information between all adjacent point data among the plurality of point data;

[0007] Group the plurality of point data according to the error information to obtain at least one data group;

[0008] According to the data information of the point data in the data group, correspondingly determine whether the point data in the data group is noise data, and remove the point data determined to be noise data.

[0009] In one embodiment, the data information includes the quantity and quality information of the point data, and the step of correspondingly determining whether the point data in the data group is noise data by using the data information of the point data in the data group includes:

[0010] When the quantity of the point data in the data group is less than a first preset quantity, correspondingly determine that the point data in the data group is the noise data;

[0011] When the quantity of the point data in the data group is equal to the first preset quantity, or greater than the first preset quantity and less than a second preset quantity, or equal to the second preset quantity, then determine whether the point data in the data group is the noise data according to the quality of the point data in the data group;

[0012] When the number of the point data in the data group is greater than the second preset number, it is correspondingly determined that the point data in the data group is valid data.

[0013] In one embodiment, the determining whether the point data in the data group is the noise data according to the quality of the point data in the data group includes:

[0014] Sum the quality of all the point data in the data group to obtain a quality sum value;

[0015] Perform an averaging process on the quality sum value according to the number of the point data in the data group to obtain a quality average value;

[0016] When the quality average value is greater than a preset quality value, it is determined that the point data in the data group is the valid data;

[0017] When the quality average value is less than or equal to the preset quality value, it is determined that the point data in the data group is the noise data.

[0018] In one embodiment, the error information groups the multiple point data to obtain at least one data group, including:

[0019] Group the multiple point data according to the angular error and / or distance error between adjacent point data to obtain at least one data group.

[0020] In one embodiment, the grouping the multiple point data according to the angular error and / or distance error between adjacent point data to obtain at least one data group includes:

[0021] Select a starting point data from the adjacent point data and classify the starting point data into the first data group;

[0022] Take the adjacent point data as the previous point data, and determine the next point data adjacent to the starting point data according to the data sequence of the adjacent point data;

[0023] Judge whether the previous point data and the next point data meet a preset condition; the preset condition includes that the angular error is less than a preset angular error value and / or the distance error is less than a preset distance error value;

[0024] When the preset condition is met, classify the next point data into the data group corresponding to the previous point data. When the preset condition is not met, create a data group corresponding to the next point data and classify the next point data into the data group corresponding to the next point data;

[0025] According to the data sequence of the adjacent point data, use the next point data as the new previous point data, and confirm the new next point data according to the new previous point data.

[0026] In one embodiment, the obtaining the target data set includes:

[0027] Obtain the original radar data detected by the radar;

[0028] According to the protocol rules corresponding to the radar, perform data parsing on the original radar data to obtain the parsed radar data, and the parsed radar data includes a plurality of point data continuously detected by the radar;

[0029] From the parsed radar data, screen out a plurality of point data obtained by the radar scanning one week for splicing processing to obtain the target data set.

[0030] In one embodiment, after determining whether the point data in the data group is noise data through the data information of the point data in the data group, the method further includes:

[0031] When there is the noise data, eliminate the noise data.

[0032] In one embodiment, after eliminating the noise data, the method further includes:

[0033] Splice all the valid data in data order to obtain the target valid data corresponding to one circle of radar scanning;

[0034] Perform object recognition according to the target valid data.

[0035] The second aspect of the present application provides a radar data processing device, and the radar data processing device includes:

[0036] An obtaining module, configured to obtain a target data set, where the target data set includes a plurality of point data continuously detected by the radar; and is further configured to obtain the error information between all adjacent point data in the plurality of point data;

[0037] A processing module, configured to group the plurality of point data according to the error information to obtain at least one data group; and is further configured to correspondingly determine whether the point data in the data group is noise data through the data information of the point data in the data group.

[0038] The third aspect of the present application provides a computer device, including a memory and a processor, where a computer program is stored in the memory, and the processor is characterized in that when the processor is used to call and execute the computer program, the steps implemented by the data processing method described in any item of the foregoing first aspect are implemented.

[0039] In one embodiment, the computer device includes a robot.

[0040] The fourth aspect of the present application provides a computer-readable storage medium storing a computer program, characterized in that the computer program, when executed by a processor, implements the steps implemented by the data processing method according to any one of the foregoing first aspects.

[0041] In one of the above-provided solutions, after obtaining the radar data, the error information between adjacent point data is grouped, and then it is determined whether the point data in the data group is noise data according to the data information in the point data of the data group, which can effectively remove the noise data in the radar data and facilitate subsequent use. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for the description of the embodiments of the present application will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0043] Figure 1 is a schematic structural diagram of a computer device in an embodiment of the present application;

[0044] Figure 2 is a schematic structural diagram of a robot in an embodiment of the present application;

[0045] Figure 3 is a schematic flowchart of a data processing method in an embodiment of the present application;

[0046] Figure 4 is a schematic diagram showing the distribution of noise data and valid data of a data processing method in an embodiment of the present application;

[0047] Figure 5 is a schematic structural diagram of a radar data processing device in an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0048] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are some, but not all, of the embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts belong to the scope of protection of the present application.

[0049] This application provides a data processing method, related devices, equipment, and media. To facilitate the understanding of this application, this application will be described one by one from the above-mentioned various topics, and will be described separately from aspects such as methods, equipment, and devices.

[0050] A. A computer device

[0051] In one embodiment, as Figure 1 shown, a computer device is provided, including a memory and a processor. The computer device can obtain radar data detected by a radar, and can perform object recognition through the radar data detected by the radar, such as methods for obstacle recognition, mapping, etc. Among them, the memory stores program code, and the transceiver is used to receive or send data / instructions. For example, the computer device uses the transceiver to receive radar data through the transceiver. The radar can be a lidar. The computer device can be a robot, a server, a terminal device, or other devices with computing capabilities and data transceiver capabilities, and is not specifically limited. The processor is used to call the program code, and when the program code is executed, the computer device is arranged to implement the following steps:

[0052] S101: Obtain a target data set, where the target data set includes multiple point data continuously detected by the radar;

[0053] S102: Obtain the error information between all adjacent point data in the multiple point data;

[0054] S103: Group the multiple point data according to the error information to obtain at least one data group;

[0055] S104: Determine whether the point data in the data group is noise data according to the data information of the point data in the data group, where the data information includes the quantity and quality information of the point data.

[0056] In step S101, the radar can send out transmission signals to the surrounding. When the transmission signals reach the surrounding objects, corresponding reflection signals will be feedback. The reflection signals contain point data, and the point data can include data information such as distance information and angle information that reflect the position and even shape of the object. When the radar rotates and scans one week, the radar can receive all the point data feedback from the one-week scan, and the received point data has a reception order, and the point data is in order. For the computer device, a target data set will be obtained, and the target data set includes multiple point data continuously detected by the radar.

[0057] In step S102, after acquiring the target data set, the computer device will acquire the error information between adjacent point data in the multiple point data. It can be understood that each point data includes data information reflecting the position and even shape of the object. Therefore, the point data fed back from different points are different, and these differences reflect the distribution of noise data. The distribution of noise data is reflected by the differences between a series of adjacent point data. In this step, the error information between all adjacent point data in the multiple point data will be acquired. Exemplarily, assuming that the target data set includes point data 1, point data 2, point data 3, ..., point data N-1 and point data N, then point data 1 and point data 2 are adjacent point data, point data 2 and point data 3 are adjacent point data, and so on, point data N-1 and point data N are adjacent point data, and this step will acquire the error information between these adjacent point data, that is, the error information between each pair of adjacent point data, including the error information between point data 1 and point data 2, ... and the error information between point data N-1 and point data N.

[0058] In step S103, after obtaining the error information between each pair of adjacent point data, the computer device will group the multiple point data according to the error information between each pair of adjacent point data to obtain at least one data group. It should be noted that the objects around the radar detection are diverse, including the position, size and shape of the objects. In this step, after obtaining the error information between each pair of adjacent point data, the multiple point data will be grouped according to the error information between each pair of adjacent point data. The error information less than the preset error will be grouped in the same data group, and the error information greater than or equal to the preset error will create a new data group for placement. It can be seen that the point data with large differences will be grouped in the same data group, thereby obtaining at least one data group. It can also be further seen that a point corresponds to one of the objects scanned by the radar. Because the error information between the point data reflected back from the same object is relatively small.

[0059] In step S104, after all adjacent point data are grouped, the data information of the point data in each data group can be determined. From the above analysis, it can be seen that the point data in a data group reflects the point data of the same object. If the point data in a data group is relatively small or the data quality does not meet the requirements, then the point data in the data group is very likely to be noise data, that is, it is not the point data reflected by the actual object. Therefore, in this step, it can be determined whether the point data in the data group is noise data according to the quantity and / or quality of the point data in the data group.

[0060] It can be seen from this embodiment that a computer device is provided. After obtaining radar data, the computer device uses the error information between adjacent point data for grouping, and then determines whether the point data in the data group is noise data based on the data information in the point data of the data group, which can effectively remove the noise data in the radar data and facilitate subsequent use.

[0061] It should be noted that, as one of the specific embodiments, a method for specifically determining whether the point data in the data group is noise data according to the data information of the point data in the data group is provided. That is, when the program code is executed, the computer device is arranged to implement the following steps specifically:

[0062] S1041: When the number of the point data in the data group is less than the first preset number, it is correspondingly determined that the point data in the data group is the noise data;

[0063] S1042: When the number of the point data in the data group is equal to the first preset number, or greater than the first preset number and less than the second preset number, or equal to the second preset number, it is determined whether the point data in the data group is the noise data according to the quality of the point data in the data group;

[0064] S1043: When the number of the point data in the data group is greater than the second preset number, it is correspondingly determined that the point data in the data group is valid data.

[0065] For steps S1041 - S1043, several determination ranges are defined for the number of point data in the data group, which are divided by using the first preset number and the second preset number. Different defined ranges have different processing methods. Among them, the first preset number and the second preset number can be an empirical value and are not specifically limited.

[0066] According to the divided number range of the point data, the following different determination methods are as follows:

[0067] Method 1: When the number of the point data in the data group is less than the first preset number, it is correspondingly determined that the point data in the data group is the noise data. Exemplarily, the first preset number can be 2. Then, when the number of point data in a certain data group 1 is less than 2, for example, the number of point data in this data group 1 is 1, it will be determined that this point data in the data group 1 is noise data. As analyzed above, there is only 1 point data. Based on the division characteristics of the data group, it shows that there are significant differences between the point data in this data group and the point data in other data groups, and the quantity is relatively small, so it can basically be considered as noise data.

[0068] Method 2: When the number of the point data in the data group is greater than the second preset number, it is correspondingly determined that the point data in the data group are valid data. Exemplarily, the second preset number can be 4. Then, when the number of the point data in a certain data group 2 is greater than 4, for example, the number of the point data in the data group 2 is 5, it indicates that there are 5 point data in the data group 2, and it will be determined that all the point data in the data group 2 are valid data, and the valid data are non-noise data. As analyzed above, when the number of the point data in a certain data group is large enough, it indicates that there are significant differences between the point data in this data group and the point data in other data groups, and a large enough number can indicate that the point data are reflected by an object and can basically be considered as valid data.

[0069] Method 3: Method 3 is different from the processing methods of Method 1 and Method 2. When the number of the point data in the data group is equal to the first preset number, or greater than the first preset number and less than the second preset number, or equal to the second preset number, the inventor has found that there will be certain errors if it is directly determined whether the data are noise data based on the number in the data group. Therefore, in this Method 3, it is possible to further determine whether the point data in the data group are the noise data according to the quality of the point data in the data group. Exemplarily, still taking the first preset number as 2 and the second preset number as 4 as an example, then when the number of the point data in a certain data group 3 is 2, or the number of the point data in the data group 3 is 4, or the number of the point data in the data group 3 is between 2 and 4, then it will be determined whether the point data in the data group 3 are noise data according to the quality of the point data in the data group 3. It should be noted that the point data can include other information in addition to the angle information and distance information, such as quality information. In this method, the quality of the point data is further utilized, and combined with the grouping processing method, it is determined whether the point data in the divided data group are noise data. Among them, the above quality information is a value calculated based on a certain quality evaluation system of the point data and is used to reflect the quality of the point data. The specific calculation process is not elaborated here in detail.

[0070] It should be noted that in this embodiment, the computer device is configured to implement the methods of determining whether the point data in each data group are noise data in the above three methods, which improves the feasibility and has a high application scenario. Among them, in Method 3, that is, in step S1042, there can be multiple ways to determine whether the point data in the data group are the noise data according to the quality of the point data in the data group. Among them, as an implementation method, that is, when the program code is executed, the computer device is arranged to further implement the following steps:

[0071] S10421: Sum the quality of all the point data in the data group to obtain a quality sum value;

[0072] S10422: Perform a mean processing on the quality sum value according to the number of the point data in the data group to obtain a quality mean value;

[0073] S10423: When the quality mean value is greater than a preset quality value, determine that the point data in the data group is the valid data;

[0074] S10424: When the quality mean value is less than or equal to the preset quality value, determine that the point data in the data group is the noise data.

[0075] For steps S10421 - S10424, a specific method for determining whether the data group is noise data based on the quality of the point data in the data group is provided. Let the quality mean value of data group i be P i , the number of the corresponding point data of this data group i be ni, and the quality sum value of the corresponding point data of this data group i be Si. Then Pi = Si / ni. Finally, by judging the situation of the quality mean value Pi and the preset quality value, the noise data situation of the point data in data group i is determined. Exemplarily, for example, the number of the point data in data group 4 is 3, and the set quality values of these three point data are a, b, and c respectively. Then the quality sum value of these three point data S4 = a + b + c, and the quality mean value P4 of data group 4 = S4 / 3. When the quality mean value P4 is greater than the preset quality value, determine that the point data in data group 4 is the valid data; when the quality mean value P4 is less than or equal to the preset quality value, determine that the point data in data group 4 is the noise data.

[0076] It should be noted that the preset quality value can be calculated according to experience or a certain calculation method. As a specific implementation manner, the preset quality value = quality coefficient * best quality value, where the quality coefficient can be set and is not specifically limited, and the best quality value can be the quality value corresponding to the point data with the highest quality among all the point data in the target data set.

[0077] In addition, it should be noted that the above quality mean value can be an arithmetic mean value or other mean values, which are not limited here.

[0078] In step S103 of the above embodiment, it is mentioned that the multiple point data will be grouped according to the error information between each pair of the adjacent point data to obtain at least one data group. Specifically, the error information between each pair of the adjacent point data can specifically be distance error and / or angle error. Therefore, as a specific implementation manner, when the program code is executed, the computer device is arranged to further implement the following steps: Group the multiple point data according to the angle error and / or distance error between each pair of the adjacent point data to obtain at least one of the data groups, that is, one of at least three data group methods can be selected to perform data grouping: The first is to group the multiple point data according to the angle error between each pair of the adjacent point data to obtain at least one of the data groups; the second is to group the multiple point data according to the distance error between each pair of the adjacent point data to obtain at least one of the data groups; the third is to group the multiple point data according to both the distance error and the angle error between each pair of the adjacent point data to obtain at least one of the data groups.

[0079] More specifically, as a specific implementation manner, when the program code is executed, the computer device is arranged to further implement the following steps:

[0080] a. Select a starting point data from the adjacent point data and classify the starting point data into the first data group;

[0081] In step a, if the target data set is the first circle of data just received when the radar is started, then directly select the first point as the starting point, and the point data corresponding to the starting point is the starting point data. For other circles of data, then directly use the data of the last point of the previous circle as the starting point data. After selecting the starting point, the starting point data is default in the first data group, that is, the first data group.

[0082] b. Take the starting point data as the previous point data and determine the next adjacent point data of the starting point data according to the data order of the adjacent point data;

[0083] c. Determine whether the previous point data and the next point data meet a preset condition; the preset condition includes that the angle error is less than a preset angle error value and / or the distance error is less than a preset distance error value;

[0084] d. When the preset condition is met, classify the next point data into the data group corresponding to the previous point data; when the preset condition is not met, create a data group corresponding to the next point data and classify the next point data into the corresponding data group;

[0085] e. According to the data sequence of the adjacent point data, take the next point data as the new previous point data, confirm the new next point data based on the new previous point data, and repeat steps c - d until the adjacent point data is grouped completely.

[0086] In steps b - e, after selecting the starting point data, according to the data sequence of the adjacent point data, sequentially determine whether the previous point data and the next point data in the adjacent point data meet the preset conditions. Here, taking the preset conditions as needing to simultaneously meet two conditions as an example: Condition 1: The angle error is less than the preset angle error value; Condition 2: The distance error is less than the preset distance error value.

[0087] Exemplarily, the adjacent point data includes point data 1, point data 2,..., point data N - 1, point data N. After selecting point data 1 as the starting point data, starting from this point data 1, select the initial previous point data and the next point data, and determine whether the two conditions are met between point data 1 (previous point data) and point data 2 (next point data). If either of the above conditions is not met, it means that the preset conditions are not met, and a new second data group will be generated, and this point data 2 will be put into the second data group; if both of the above conditions are not met, then this point data 2 will be put into the first data group; then according to the data sequence, continuously select the new next point data to compare with the new previous point data, that is, continue to select point data 2 and point data 3 for comparison. At this time, point data 2 is the previous point data, and point data 3 is the next point data. After the comparison between point data 2 and point data 3, continue to select point data 3 and point data 4 for comparison until point data N - 1 and point data N are compared. That is, continuously compare the next point data with the previous point data, and repeat the above process until all the point data in the adjacent point data are traversed. At this time, the point data in the target data set has been grouped, and at least one data group is obtained.

[0088] Exemplarily, for the grouping result, the situations of noise data and valid data can be as Figure 4 shown.

[0089] Among them, it should be noted that, as an implementation manner, the preset angle error value can be (sin -1 0.005 / min(x, y)), where x represents the distance of the previous point data, and y represents the distance of the next point data. As an implementation manner, the preset distance error value can be 0.0005 meters. It should be noted that the above example is only an exemplary illustration and is not specifically limited.

[0090] In addition, it should be noted that the preset conditions can also be to meet one of the conditions, without specific limitations and no further elaboration. In this embodiment, various specific processing methods for data groups are proposed, which improves the feasibility of the solution.

[0091] As described above, when the radar rotates and scans a full circle, the radar can receive all the point data feedback from the scan of a full circle, and the received point data has a reception order, and the data is in order. For a computer device, a target data set will be obtained, and the target data set includes a plurality of point data continuously detected by the radar. In this embodiment, since relevant information of the point data will be used later, the obtained target data set can be the target data set reflected after the radar is parsed, or the target data set parsed by the computer device. That is, in one embodiment, when the program code is executed, the computer device is arranged to specifically implement the following steps:

[0092] S1011: Obtain the original radar data detected by the radar;

[0093] S1012: According to the protocol rules corresponding to the radar, perform data parsing on the original radar data to obtain the parsed radar data, and the parsed radar data includes a plurality of point data continuously detected by the radar;

[0094] S1013: From the parsed radar data, screen out a plurality of point data obtained by the radar scanning a full circle for splicing processing to obtain the target data set.

[0095] For steps S1011 - S1013, first, the computer device will receive the original radar data sent by the radar, and then, according to the protocol rules corresponding to the radar, that is, according to the protocol rules given by the manufacturer, perform parsing on the radar data obtained by the radar to obtain the parsed radar data. It can be understood that the radar data protocols of different manufacturers and / or different models are different. Therefore, it is necessary to parse the data according to the manufacturer's protocol rules, and then perform splicing processing on the parsed data according to the rules of scanning a full circle, and merge the individual point data into the data of scanning a full circle. Specifically, the rule for judging the end of a circle can be that the angle of the point data received later is smaller than the angle of the previous point data. In this way, after parsing, the point data parsed by the computer device can include the angle and distance information of the point data, which is convenient for the subsequent implementation of the solution. All the point data feedback when the radar rotates a full circle is the target data set to be spliced. The splicing rule is that for each received point data, it is judged whether it is smaller than the angle of the previous point. If it is smaller, it is considered as the starting data of the next circle of scanning, which is convenient for finally outputting to obtain the target data set.

[0096] In this embodiment, the processed radar data will be spliced to obtain a target data set corresponding to each scan of the radar. Then, subsequent noise processing will be performed based on the point data set of each week. On the one hand, it makes the data processing standardized. On the other hand, due to the data difference problems in different scan cycles, this classification is also beneficial to screening out the real noise data.

[0097] It should be noted that in the computer device provided in the above embodiment, finally, the noise data in the point data corresponding to one scan of the radar will be determined. As an embodiment, when the program code is executed, after determining whether the point data in the data group is noise data through the number of the point data in the data group, the computer device is arranged to further implement the following steps:

[0098] S105: When there is such noise data, the noise data is removed.

[0099] S106: Splice all the valid data in the data order to obtain the target valid data corresponding to one scan of the radar;

[0100] S107: Perform object recognition based on the target valid data.

[0101] In step S105, in this embodiment, after the computer device determines the noise data, the noise data can be removed. In this way, for the computer device, through the computer device provided by the embodiment of the present application, the valid data corresponding to the weekly radar data can be obtained, which is beneficial to the subsequent computer device to carry out various data application works based on the valid radar data.

[0102] For steps S106 - S107, a data application method is provided. The computer device will splice all the valid data in the data order to obtain the target valid data corresponding to one scan of the radar, and then perform object recognition according to the target valid data. Through this embodiment, the computer device can perform object recognition based on the valid radar data. Since the noise data is removed, the influence of the noise on the radar scan data is eliminated, and the accuracy of object recognition can be effectively improved, and the object scanned by the radar can be accurately restored.

[0103] It should be noted that the above embodiment is only taken object recognition as an example. Based on object recognition, the computer device can also perform subsequent map building and other works, which is beneficial to improving the map building accuracy, and will not be elaborated here one by one.

[0104] B. A robot

[0105] A robot is provided, such as Figure 2As shown, it includes a radar, a memory, and a processor. The radar can be a radar device such as a lidar. The radar is deployed on a robot. The robot can perform object recognition, such as obstacle recognition and mapping, based on the radar data detected by the radar. Among them, the memory stores program code, and the transceiver is used to receive or send data / instructions. For example, the robot uses the transceiver to receive the data fed back by the radar. The processor is used to call the program code, and when the program code is executed, the robot is arranged to implement the following steps:

[0106] S201: Obtain a target data set, where the target data set includes a plurality of point data continuously detected by the radar;

[0107] S202: Obtain the error information between all adjacent point data among the plurality of point data;

[0108] S203: Group the plurality of point data according to the error information to obtain at least one data group;

[0109] S204: Determine whether the point data in the data group is noise data according to the data information of the point data in the data group.

[0110] Among them, steps S201 - S204 can refer to steps S101 - 104 in the foregoing embodiments correspondingly. However, the acting subject of the embodiments of this application is specifically a robot, and the radar is deployed on the robot. The robot can be a food delivery robot, a mapping robot, a navigation robot, a guiding robot, etc., and is not specifically limited.

[0111] It can be seen from this embodiment that a specific robot is provided. After the robot obtains radar data, it groups the error information between adjacent point data, and then determines whether the point data in the data group is noise data according to the number of point data in the data group. It can effectively remove the noise data in the radar data, facilitate the subsequent use of the robot, including object recognition, etc., and improve the accuracy of the robot's object recognition.

[0112] It should be noted that for more steps or functions that the robot can implement, reference can be made to the descriptions of the various embodiments of the foregoing computer device, and they will not be repeated here one by one.

[0113] C. A data processing method

[0114] The above describes the computing device and the robot. As Figure 3 shown, the embodiments of this application also provide a data processing method, including the following steps:

[0115] S301: Obtain a target data set, where the target data set includes a plurality of point data continuously detected by a radar;

[0116] S302: Obtain the error information between all adjacent point data among the plurality of point data;

[0117] S303: Group the plurality of point data according to the error information to obtain at least one data group;

[0118] S304: Determine whether the point data in the data group is noise data according to the data information of the point data in the data group.

[0119] Among them, the content of the above steps S301 - S304 can refer to the corresponding content of steps S101 - S104 in the foregoing embodiment, and will not be described repeatedly here.

[0120] It can be seen from this embodiment that a data processing method is provided. After obtaining radar data, the error information between adjacent point data is used for grouping, and then whether the point data in the data group is noise data is determined according to the quantity of the point data in the data group, which can effectively remove the noise data in the radar data and facilitate subsequent use.

[0121] It should be noted that as one specific embodiment, a specific method for determining whether the point data in the data group is noise data according to the quantity of the point data in the data group is provided. That is, in step S304, that is, determining whether the point data in the data group is noise data according to the quantity of the point data in the data group specifically includes the following steps:

[0122] S3041: When the quantity of the point data in the data group is less than a first preset quantity, it is correspondingly determined that the point data in the data group is the noise data;

[0123] S3042: When the quantity of the point data in the data group is equal to the first preset quantity, or greater than the first preset quantity and less than a second preset quantity, or equal to the second preset quantity, it is determined whether the point data in the data group is the noise data according to the quality of the point data in the data group;

[0124] S3043: When the quantity of the point data in the data group is greater than the second preset quantity, it is correspondingly determined that the point data in the data group is valid data.

[0125] For steps S3041 - S3043, reference can be made to the relevant descriptions in the foregoing embodiment S1041 - S1043, and will not be described repeatedly here.

[0126] In one embodiment, that is, in step S3042, there are also various ways to determine whether the point data in the data group is the noise data according to the quality of the point data in the data group. One of the ways specifically includes the following steps:

[0127] S30421: Sum the quality of all the point data in the data group to obtain a quality sum value;

[0128] S30422: Perform an averaging process on the quality sum value according to the number of the point data in the data group to obtain a quality average value;

[0129] S30423: When the quality average value is greater than a preset quality value, determine that the point data in the data group is the valid data;

[0130] S30424: When the quality average value is less than or equal to the preset quality value, determine that the point data in the data group is the noise data.

[0131] For steps S30421 - S30424, reference can be made to the relevant descriptions of steps S10421 - S10424 in the foregoing embodiment, and the description will not be repeated here.

[0132] In one embodiment, in step S303 of the above embodiment, that is, grouping the multiple point data according to the error information between each pair of adjacent point data to obtain at least one data group, the specific steps are as follows: Group the multiple point data according to the angular error and / or distance error between each pair of adjacent point data to obtain at least one data group.

[0133] More specifically, as a specific implementation manner, the above S303 specifically includes the following steps:

[0134] a. Select a starting point data from the adjacent point data and classify the starting point data into the first data group;

[0135] b. Take the starting point data as the previous point data and determine the next adjacent point data of the starting point data according to the data sequence of the adjacent point data;

[0136] c. Determine whether the previous point data and the next point data meet a preset condition; the preset condition includes that the angular error is less than a preset angular error value and / or the distance error is less than a preset distance error value;

[0137] d. When the preset condition is met, the next point data is classified into the data group corresponding to the previous point data; when the preset condition is not met, a data group corresponding to the next point data is created and the next point data is classified into the data group corresponding to the next point data;

[0138] e. According to the data sequence of the adjacent point data, the next point data is used as the new previous point data, and the new next point data is confirmed according to the new previous point data, and steps cd are repeated until the adjacent point data are completely grouped.

[0139] For explanations about steps ae, please refer to the corresponding description in the aforementioned computer device embodiment, which will not be repeated here.

[0140] In one embodiment, step S301, that is, obtaining a target data set, specifically includes the following steps:

[0141] S3011: Acquire raw radar data detected by the radar;

[0142] S3012: parsing the original radar data according to a protocol rule corresponding to the radar to obtain parsed radar data, where the parsed radar data includes a plurality of point data continuously detected by the radar;

[0143] S3013: Filter out a plurality of point data obtained by scanning one circle of the radar from the analyzed radar data, and perform splicing processing to obtain the target data set.

[0144] For steps S3011 - S3013 , reference may be made to the relevant descriptions of steps S3011 - S3013 in the aforementioned embodiments, and the descriptions are not repeated here.

[0145] In one embodiment, after step S304, that is, after determining whether the point data in the data group is noise data through the data information of the point data in the data group, the method further includes the following steps:

[0146] S305: When the noise data exists, remove the noise data.

[0147] S306: All valid data are spliced in data order to obtain valid target data corresponding to one round of radar scanning;

[0148] S307: Perform object recognition according to the target valid data.

[0149] For steps S305-S307, reference may be made to the relevant description of steps S105-S107 in the aforementioned embodiment, and the description will not be repeated here.

[0150] It should be noted that the above embodiments only take object recognition as an example, for basic object recognition. This data processing method can also perform subsequent mapping and other tasks, which is beneficial to improving the mapping accuracy and will not be elaborated here one by one.

[0151] It should be understood that the magnitudes of the sequence numbers of the steps in the above embodiments do not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present application.

[0152] D. A radar data processing device

[0153] In one embodiment, a radar data processing device is provided, and this radar data processing device corresponds one by one to the data processing method in the above embodiment. As Figure 5 shown, the radar data processing device 10 includes an acquisition module 101 and a processing module 102. The detailed description of each functional module is as follows:

[0154] The acquisition module 101 is used to acquire a target data set, and the target data set includes a plurality of point data continuously detected by the radar; it is also used to acquire the error information between all adjacent point data in the plurality of point data;

[0155] The processing module 102 is used to group the plurality of point data according to the error information to obtain at least one data group; it is also used to correspondingly determine whether the point data in the data group is noise data through the data information of the point data in the data group.

[0156] It can be seen from this embodiment that a radar data processing device is provided. After acquiring radar data, it groups the error information between adjacent point data, and then determines whether the point data in the data group is noise data according to the number of point data in the data group, which can effectively remove the noise data in the radar data and facilitate subsequent use.

[0157] In one embodiment, the processing module 102 is further specifically used for:

[0158] When the number of the point data in the data group is less than a first preset number, it is correspondingly determined that the point data in the data group is the noise data;

[0159] When the number of the point data in the data group is equal to the first preset number, or greater than the first preset number and less than a second preset number, or equal to the second preset number, it is correspondingly determined whether the point data in the data group is the noise data according to the quality of the point data in the data group;

[0160] When the number of the point data in the data group is greater than the second preset number, it is correspondingly determined that the point data in the data group is valid data.

[0161] In one embodiment, the processing module 102 is configured to correspondingly determine whether the point data in the data group is the noise data according to the quality of the point data in the data group, including:

[0162] The processing module 102 is specifically configured to:

[0163] Sum the quality of all the point data in the data group to obtain a quality sum value;

[0164] Perform an averaging process on the quality sum value according to the number of the point data in the data group to obtain a quality average value;

[0165] When the quality average value is greater than a preset quality value, it is determined that the point data in the data group is the valid data;

[0166] When the quality average value is less than or equal to the preset quality value, it is determined that the point data in the data group is the noise data.

[0167] In one embodiment, the processing module 102 is further specifically configured to: group the multiple point data according to the angular error and / or distance error between each pair of adjacent point data to obtain at least one data group.

[0168] In one embodiment, the processing module 102 is further specifically configured to:

[0169] a. Select a starting point data from the adjacent point data and classify the starting point data into a first data group;

[0170] b. Take the starting point data as the previous point data and determine the next point data adjacent to the starting point data according to the data order of the adjacent point data;

[0171] c. Determine whether the previous point data and the next point data meet a preset condition; the preset condition includes that the angular error is less than a preset angular error value and / or the distance error is less than a preset distance error value;

[0172] d. When the preset condition is met, classify the next point data into the data group corresponding to the previous point data; when the preset condition is not met, create a data group corresponding to the next point data and classify the next point data into the data group corresponding to the next point data;

[0173] e. According to the data order of the adjacent point data, use the next point data as the new previous point data, confirm the new next point data based on the new previous point data, and repeat steps c - d until the adjacent point data is completely grouped.

[0174] In one embodiment, the obtaining module 101 is specifically configured to:

[0175] Obtain the original radar data detected by the radar;

[0176] According to the protocol rules corresponding to the radar, perform data parsing on the original radar data to obtain the parsed radar data, and the parsed radar data includes multiple point data continuously detected by the radar;

[0177] From the parsed radar data, screen out multiple point data obtained by the radar scanning one week for splicing processing to obtain the target data set.

[0178] In one embodiment, the processing module 102 is further specifically configured to: when there is the noise point data, eliminate the noise point data.

[0179] In one embodiment, the processing module 102 is further specifically configured to: after eliminating the noise point data, splice all the valid data in data order to obtain the target valid data corresponding to one circle of radar scanning; perform object recognition based on the target valid data.

[0180] For the specific limitations of the radar data processing device, reference can be made to the relevant limitations of the computer device in the foregoing embodiments, which will not be elaborated here. Each module in the above - mentioned radar data processing device can be implemented in whole or in part through software, hardware, and their combination. The above - mentioned modules can be embedded in the processor of the computer device in hardware form or be independent of it, or be stored in the memory of the computer device in software form, so that the processor can call and execute the operations corresponding to the above - mentioned modules.

[0181] E. A computer - readable storage medium

[0182] In one embodiment, a computer - readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, it implements the steps of the data processing method in the foregoing embodiments, or implements the functions of the radar data processing device in the foregoing embodiments, or implements the functions of the computer device in the foregoing embodiments, or implements the functions of the robot in the foregoing embodiments.

[0183] Among them, for more steps or functions implemented when the computer program is executed by the processor, reference can be made to the relevant descriptions in the foregoing embodiments, and no repeated description will be given here.

[0184] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, storage, database, or other medium used in the embodiments provided in the present application can include non-volatile and / or volatile memories. Non-volatile memories can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memories can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and Rambus dynamic RAM (RDRAM), etc.

[0185] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the above division of each functional unit and module is used as an example. In actual applications, the above functions can be allocated to different functional units and modules according to needs, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0186] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the 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, and should all be included in the protection scope of the present application.

Claims

1. A data processing method, characterized in that, The method includes: Obtaining a target data set, where the target data set includes a plurality of point data continuously detected by a radar; Obtaining error information between all adjacent point data among the plurality of point data; Grouping the plurality of point data according to the error information to obtain at least one data group; Correspondingly determining whether the point data in the data group is noise data according to the data information of the point data in the data group; Removing the point data determined to be noise data; The data information includes the quantity and quality information of the point data. Correspondingly determining whether the point data in the data group is noise data according to the data information of the point data in the data group includes: When the quantity of the point data in the data group is less than a first preset quantity, correspondingly determining the point data in the data group as the noise data; When the quantity of the point data in the data group is equal to the first preset quantity, or greater than the first preset quantity and less than a second preset quantity, or equal to the second preset quantity, then sum the quality of all the point data in the data group to obtain a quality sum value; perform an averaging process on the quality sum value according to the quantity of the point data in the data group to obtain a quality average value; when the quality average value is greater than a preset quality value, determine that the point data in the data group is valid data; when the quality average value is less than or equal to the preset quality value, determine that the point data in the data group is the noise data; When the quantity of the point data in the data group is greater than the second preset quantity, correspondingly determining the point data in the data group as valid data.

2. The data processing method according to claim 1, wherein The grouping the plurality of point data according to the error information to obtain at least one data group includes: Grouping the plurality of point data according to the angular error and / or distance error between adjacent point data to obtain at least one data group.

3. The data processing method according to claim 2, wherein The grouping the plurality of point data according to the angular error and / or distance error between adjacent point data to obtain at least one data group includes: Selecting a starting point data from the adjacent point data and classifying the starting point data into a first data group; Taking the starting point data as the previous point data and determining the next adjacent point data of the starting point data according to the data order of the adjacent point data; Judging whether the previous point data and the next point data meet a preset condition; the preset condition includes that the angular error is less than a preset angular error value and / or the distance error is less than a preset distance error value; When the preset condition is met, classifying the next point data into the data group corresponding to the previous point data; when the preset condition is not met, creating a data group corresponding to the next point data and classifying the next point data into the data group corresponding to the next point data; According to the data order of the adjacent point data, taking the next point data as the new previous point data and determining the new next point data according to the new previous point data.

4. The data processing method according to claim 1, wherein The obtaining the target data set includes: Obtain the original radar data detected by the radar; According to the protocol rules corresponding to the radar, perform data parsing on the original radar data to obtain the parsed radar data, where the parsed radar data includes multiple point data continuously detected by the radar; From the parsed radar data, screen out multiple point data obtained by the radar scanning one week for splicing processing to obtain the target data set.

5. The data processing method according to claim 1, wherein After determining whether the point data in the data group is noise data according to the data information of the point data in the data group, the method further includes: Splice all valid data in data order to obtain the target valid data corresponding to one circle of radar scanning; Perform object recognition according to the target valid data.

6. A radar data processing device, characterized in that, The radar data processing device includes: An acquisition module, configured to acquire a target data set, where the target data set includes multiple point data continuously detected by the radar; and is further configured to acquire the error information between all adjacent point data in the multiple point data; A processing module, configured to group the multiple point data according to the error information to obtain at least one data group; and is further configured to correspondingly determine whether the point data in the data group is noise data according to the data information of the point data in the data group; The data information includes the quantity and quality information of the point data. Corresponding determination of whether the point data in the data group is noise data according to the data information of the point data in the data group includes: when the quantity of the point data in the data group is less than a first preset quantity, correspondingly determine that the point data in the data group is the noise data; when the quantity of the point data in the data group is equal to the first preset quantity, or greater than the first preset quantity and less than a second preset quantity, or equal to the second preset quantity, sum the quality of all the point data in the data group to obtain a quality sum value; perform mean processing on the quality sum value according to the quantity of the point data in the data group to obtain a quality mean value; when the quality mean value is greater than a preset quality value, determine that the point data in the data group is valid data; when the quality mean value is less than or equal to the preset quality value, determine that the point data in the data group is the noise data; when the quantity of the point data in the data group is greater than the second preset quantity, correspondingly determine that the point data in the data group is valid data.

7. A computer device, characterized in that, It includes a memory and a processor, and a computer program is stored in the memory. It is characterized in that when the processor is used to call and execute the computer program, the steps implemented by the data processing method according to any one of claims 1-5 are realized.

8. A computer-readable storage medium storing a computer program, characterized in that, The computer program is used to realize the steps implemented by the data processing method according to any one of claims 1-5 when executed by the processor.

Images