A method, device, equipment and storage medium for processing critical noise data

Abstract

The embodiment of the present application discloses a method and apparatus, device, and storage medium for processing critical noise data, wherein the method includes: obtaining critical noise data of the surface of the object to be detected; smoothing the critical noise data to obtain smoothed critical noise data; determining the smoothed critical noise data that is not within the first data range as defect data, wherein the first data range refers to the data range set according to the standard value and the design tolerance; removing the defect data from the smoothed critical noise data to obtain filtered critical noise data; and using the filtered critical noise data to detect the surface of the object to be detected. In this way, by processing, screening, and troubleshooting the critical noise data, more critical noise data can be placed within the data standard data range, reducing misjudgment of critical noise data, and solving the problem of over-killing critical noise data.

Description

Technical Field

[0001] The embodiments of the present application relate to but are not limited to electronic technology, and in particular to a method and apparatus, device, and storage medium for processing critical noise data. Background Art

[0002] In existing defect detection methods, a 3D ranging sensor is usually used to obtain point cloud data of the component to be detected, and the obtained point cloud data is quantitatively analyzed to complete the defect detection of the component to be detected.

[0003] When performing quantitative analysis on point cloud data, noise data is usually processed. The process of processing noise data is as follows: for obvious noise appearing in the point cloud data, it is necessary to remove the obvious noise data; and for data in the critical range of noise, that is, the value at the critical edge, since it shows the characteristics of the component to be detected itself, it is partially removed and partially retained.

[0004] However, in actual production, there are very few problems caused by production, such as aging of production tools, or short interruptions of production tools due to failures. These problems will cause some real data to be misjudged. For example, in the process of analyzing point cloud data using the critical noise processing method, the value of the real critical edge is discarded, resulting in data loss, which in turn causes the retained point cloud data to be unable to display the original physical characteristics of the defects / blemishes, which will have an irreversible impact on the subsequent result analysis. Summary of the invention

[0005] In view of this, embodiments of the present application provide a method and apparatus, device, and storage medium for processing critical noise data to solve at least one problem existing in the related art.

[0006] The technical solution of the embodiment of the present application is implemented as follows:

[0007] In a first aspect, an embodiment of the present application provides a method for processing critical noise data, the method comprising:

[0008] Acquire critical noise point data of the surface of the object to be detected;

[0009] Smoothing the critical noise point data to obtain smoothed critical noise point data;

[0010] Determine the smoothed critical noise point data that is not within a first data range as defect data, wherein the first data range refers to a data range set according to a standard value and a design tolerance;

[0011] Removing the defect data from the smoothed critical noise point data to obtain filtered critical noise point data;

[0012] The filtered critical noise point data is used to detect the surface of the object to be detected.

[0013] In a second aspect, an embodiment of the present application provides a device for processing critical noise data, the device comprising:

[0014] An acquisition module, used to acquire critical noise point data on the surface of the object to be detected;

[0015] A processing module, used for smoothing the critical noise point data to obtain smoothed critical noise point data;

[0016] A determination module, configured to determine the smoothed critical noise point data that is not within a first data range as defective data, wherein the first data range refers to a data range set according to a standard value and a design tolerance;

[0017] A filtering module, used to remove the defect data from the smoothed critical noise data to obtain filtered critical noise data;

[0018] A detection module is used to detect the surface of the object to be detected using the filtered critical noise point data.

[0019] In a third aspect, an embodiment of the present application provides a critical noise data processing device, including a memory and a processor, wherein the memory stores a computer program that can be run on the processor, and the processor implements the steps in the above-mentioned test and evidence collection method when executing the program.

[0020] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium having a computer program stored thereon, which implements the steps in the above-mentioned testing and forensics method when executed by a processor.

[0021] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0022] In an embodiment of the present application, first, the critical noise data is smoothed, and then the defective data in the smoothed critical noise data is determined. Finally, the defective data is removed from the smoothed critical noise data to obtain filtered critical noise data. In this way, by processing, screening and troubleshooting the critical noise data, more critical noise data can be made within the data standard data range, reducing the misjudgment of critical noise data, solving the problem of over-killing of critical noise data, and effectively improving the problems of critical noise data loss and small amount of critical noise data. In addition, the idea of ​​smoothing and then processing is also provided for data processing problems. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1A flowchart of a method for processing critical noise data provided in an embodiment of the present application;

[0024] Figure 2 A flowchart of a method for processing critical noise data provided in an embodiment of the present application;

[0025] Figure 3 A flowchart of a method for processing critical noise data provided in an embodiment of the present application;

[0026] Figure 4 A flowchart of a method for processing critical noise data provided in an embodiment of the present application;

[0027] Figure 5A A flowchart of a method for processing critical noise data provided in an embodiment of the present application;

[0028] Figure 5B A schematic diagram of a plane point cloud data of a product appearance surface provided in an embodiment of the present application;

[0029] Figure 5C A schematic diagram of the result of data smoothing processing of critical noise point data provided in an embodiment of the present application;

[0030] Figure 6 A schematic diagram of the composition structure of a device for processing critical noise data provided in an embodiment of the present application;

[0031] Figure 7 A hardware entity schematic diagram of a critical noise data processing device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0032] In order to better understand the embodiments of the present application, some related terms are introduced below.

[0033] 3D ranging sensor: The most commonly used method in passive ranging sensing is binocular stereo vision. This method uses two cameras separated by a certain distance to simultaneously obtain two images of the same scene, find the corresponding pixels in the two images through a stereo matching algorithm, and then calculate the disparity information based on the triangulation principle. The disparity information can be used to represent the depth information of objects in the scene through conversion.

[0034] Point cloud: In reverse engineering, the point data set of the product appearance surface obtained by measuring instruments is also called point cloud. Usually, the number of points obtained by using a 3D coordinate measuring machine is relatively small, and the distance between points is relatively large, which is called a sparse point cloud; while the point cloud obtained by using a 3D laser scanner or a photographic scanner has a relatively large number of points and is relatively dense, which is called a dense point cloud.

[0035] Smoothing: Reassigning probabilities so that even events that did not occur are given a probability.

[0036] Noise: The rough part of an image produced when light is received as a signal and outputted. It also refers to foreign pixels that should not appear in the image.

[0037] Weighted average method: The weighted average method uses several past observations of the same variable arranged in chronological order and uses the chronological number as the weight to calculate the weighted arithmetic mean of the observations. This number is used as a trend forecasting method to predict the value of the variable in the future.

[0038] The technical solution of the present application is further described in detail below with reference to the accompanying drawings and embodiments.

[0039] Figure 1 A flowchart of a method for processing critical noise data provided in an embodiment of the present application is shown in FIG. Figure 1 As shown, the method is applied to a processing device for critical noise data, such as a computer device, and includes:

[0040] Step 101: Acquire critical noise point data on the surface of the object to be detected;

[0041] Here, the surface of the object to be detected may be a product appearance surface or a part plane surface, for example, a car shell, a mobile phone screen, and a tablet screen.

[0042] Here, the critical noise data refers to data within the critical range after filtering out the noise data. The noise data refers to plane point cloud data that meets the filtering conditions. Here, the plane point cloud data refers to a set of surface point data obtained after linear scanning of the object to be detected. Here, the critical range and filtering conditions can be set according to different application scenarios.

[0043] In the implementation process, a surface point data set of the object to be detected is obtained by linearly scanning the object to be detected, that is, the plane point cloud data of the object to be detected. The filtering conditions and critical range are set according to the object to be detected, and the plane point cloud data that meets the filtering conditions is the noise data. The noise data is filtered to obtain valid data. After the valid data is screened according to the set critical range, the data within the critical range is the critical noise data.

[0044] Step 102: Smoothing the critical noise point data to obtain smoothed critical noise point data;

[0045] Here, the smoothing is to redistribute the probability, and even if an event does not occur, a probability will be assigned, which can be achieved by a smoothing algorithm. Here, the smoothing process refers to the calculation of the value of the critical noise data by using a smoothing algorithm. Through smoothing, the difference between the peak and valley values ​​of the critical noise data can be reduced, and the data fluctuation range can be reduced and the data distribution can be more concentrated under the premise of ensuring that the data size trend does not change.

[0046] Step 103: Determine the smoothed critical noise point data that is not within the first data range as defect data, wherein the first data range refers to a data range set according to a standard value and a design tolerance;

[0047] Here, the first data range can also be called the standard data range. Data that meets the standard data range is reserved data, and data that is not within the range is defective data. The first data range is determined by the standard value and the design tolerance. For example, if the standard value is set to 0.50 and the design tolerance is ±0.020, the first data range or standard data range is (0.48, 0.52).

[0048] Step 104: removing the defect data from the smoothed critical noise data to obtain filtered critical noise data;

[0049] After the defect data in the smoothed critical noise point data is determined, the defect data needs to be filtered out to obtain filtered critical noise point data for analyzing and detecting the surface of the object to be detected.

[0050] Step 105: Detect the surface of the object to be detected using the filtered critical noise point data.

[0051] In an embodiment of the present application, first, the critical noise data is smoothed, and then the defective data in the smoothed critical noise data is determined. Finally, the defective data is removed from the smoothed critical noise data to obtain filtered critical noise data. In this way, by processing, screening and troubleshooting the critical noise data, more critical noise data can be made within the data standard data range, reducing the misjudgment of critical noise data, solving the problem of over-killing of critical noise data, and effectively improving the problems of critical noise data loss and small amount of critical noise data. In addition, the idea of ​​smoothing and then processing is also provided for data processing problems.

[0052] Figure 2 A flowchart of a method for processing critical noise data provided in an embodiment of the present application is shown in FIG. Figure 2 As shown, a processing device applied to critical noise data, the method includes:

[0053] Step 201: Acquire plane point cloud data of the surface of the object to be detected;

[0054] During the implementation process, the surface of the object to be inspected is scanned by measuring instruments such as sensors, and the scanning results are transmitted to a computer device. By processing the scanning results by the computer device, a surface point data set of the object to be inspected can be obtained, a dot matrix can be formed, and plane point cloud data can be obtained.

[0055] Step 202: determining the plane point cloud data that is not within a second data range as noise data, wherein the second data range is determined by a standard value and a design tolerance; and the second data range is larger than the first data range;

[0056] Here, the second data range is a data range determined according to the standard value and the design tolerance, and is used for the first screening of the plane point cloud data. By comparing the plane point cloud data with the second data range, obvious noise data can be screened out.

[0057] It should be noted that the second data range is larger than the first data range. It can be seen that after the first screening of the second data range, the first type of noise data, that is, obvious noise data, can be removed. At this time, the defective data can be removed by screening the first data range again.

[0058] For example, by comparing with the standard value and combining the design tolerance of the part itself, the first type of noise data is removed. For example, if the standard value is 0.50 and the design tolerance is ±0.020, when the value is greater than 0.52 or less than 0.48, it is considered as the first type of noise data.

[0059] Step 203: filtering out the noise data from the plane point cloud data to obtain first point cloud data;

[0060] Here, the first point cloud data is the plane point cloud data remaining after being screened in the second data range, that is, after filtering out the obvious noise data.

[0061] Step 204: determining the first point cloud data satisfying a third data range as critical noise point data, wherein the third data range is determined by a standard value, a design tolerance, and a critical value tolerance;

[0062] Here, the third data range is a data range determined according to the standard value, the design tolerance and the critical value tolerance. For example, the standard value is 0.50, the design tolerance is ±0.020, and the critical value tolerance is: ±0.015, then the third data range is (0.48, 0.485) or (0.515, 0.52).

[0063] In one embodiment, the value of the first point cloud data is compared with the third data range, and when the value is between (0.48, 0.485) or (0.515, 0.52), the data is determined to be critical noise point data.

[0064] Step 205: Smoothing the critical noise point data to obtain smoothed critical noise point data;

[0065] Step 206: Determine the smoothed critical noise point data that is not within the first data range as defect data, wherein the first data range refers to a data range set according to a standard value and a design tolerance;

[0066] Step 207: removing the defect data from the smoothed critical noise data to obtain filtered critical noise data;

[0067] Step 208: Detect the surface of the object to be detected using the filtered critical noise point data.

[0068] In the embodiment of the present application, first, the plane point cloud data that is not within the second data range is determined as noise data. Then, the noise data is filtered out from the plane point cloud data to obtain the first point cloud data. Finally, the first point cloud data that meets the third data range is determined as critical noise data. In this way, the noise data of the plane point cloud data can be classified and screened, which is helpful for processing different types of data, optimizes the method of processing noise data, and provides a classification idea for data processing problems.

[0069] Figure 3 A flowchart of a method for processing critical noise data provided in an embodiment of the present application is shown in FIG. Figure 3 As shown, a processing device applied to critical noise data, the method includes:

[0070] Step 301: Acquire at least two different images of the surface of the object to be detected in the same scene;

[0071] Here, the at least two different images of the same scene acquired by the device are acquired by a sensor. The sensor may be a 3D ranging sensor.

[0072] Step 302: performing stereo matching algorithm on the at least two different images to obtain corresponding pixel points in the at least two different images;

[0073] Here, the stereo matching algorithm refers to an algorithm for obtaining corresponding pixel points in at least two images through matching cost calculation, cost aggregation, disparity calculation and post-processing. The stereo matching algorithm can find corresponding pixel points in at least two different images.

[0074] Step 303: Calculate the pixel value of the pixel point according to the triangulation principle to obtain the disparity information corresponding to the pixel value;

[0075] Here, the triangulation principle is to obtain disparity data such as pixel and image plane matrix, projection relationship matrix, etc. based on the consistency of the main point pixel coordinates of the two images and the triangle similarity principle.

[0076] Step 304: converting the disparity information into plane point cloud data according to a specific corresponding relationship, wherein the specific corresponding relationship is used to characterize the corresponding relationship between the disparity information and the plane point cloud data;

[0077] Step 305: Determine the plane point cloud data that is not within a second data range as noise data, wherein the second data range is determined by a standard value and a design tolerance; and the second data range is larger than the first data range;

[0078] Step 306: filtering out the noise point data from the plane point cloud data to obtain the critical noise point data;

[0079] Step 307: Smoothing the critical noise point data to obtain smoothed critical noise point data;

[0080] Step 308: Determine the smoothed critical noise point data that is not within the first data range as defect data, wherein the first data range refers to a data range set according to a standard value and a design tolerance;

[0081] Step 309: removing the defect data from the smoothed critical noise data to obtain filtered critical noise data;

[0082] Step 310: Detect the surface of the object to be detected using the filtered critical noise point data.

[0083] The present application embodiment provides a method for processing critical noise data, which is applied to a device for processing critical noise data. The method includes:

[0084] Step A1: obtaining critical noise point data of the surface of the object to be detected, wherein the critical noise point data includes a plurality of critical noise points;

[0085] Step A2: determining a weight sequence with a specific rule for each critical noise point, wherein the sequence is used to assign weights to the critical noise points in the critical noise point data;

[0086] Here, it should be noted that the device needs to use a set of weight series with a specific pattern to smooth each critical noise point, and the weights in this set of weight series are used to assign the critical noise points in the critical noise data. For example, when there are n critical noise points in the critical noise data, n is a positive integer greater than or equal to 1; there are n weight series, each weight series has n weights, and these n weights are used to assign n critical noise points.

[0087] For example, if there are 9 critical noise points in the critical noise point data to be smoothed, there are 9 weight series. For a certain critical noise point, when determining the weighted arithmetic mean of the critical noise point, 9 weights in the corresponding weight series need to be assigned to the 9 critical noise points.

[0088] Step A3: determining a weighted arithmetic mean of the corresponding critical noise point according to the weight value series and the critical noise point data, and obtaining the smoothed critical noise point data;

[0089] Here, the weighted arithmetic mean is a value obtained by multiplying each critical noise point by a corresponding weight, then summing up the total value to obtain an overall value, and then dividing the total weight by the total value.

[0090] For example, the critical noise point data that needs to be smoothed includes 9 critical noise points. For the first critical noise point y[0], the corresponding set of weight series is {17, 14, 11, 8, 5, 2, -1, -4, -7}; then the first critical noise point yy[0] after smoothing is determined based on the weight series {17, 14, 11, 8, 5, 2, -1, -4, -7} and the 9 critical noise points.

[0091] Step A4: determining the smoothed critical noise point data that is not within the first data range as defect data, wherein the first data range refers to a data range set according to a standard value and a design tolerance;

[0092] Step A5: removing the defect data from the smoothed critical noise data to obtain filtered critical noise data;

[0093] Step A6: using the filtered critical noise point data to detect the surface of the object to be detected;

[0094] Figure 4 A flowchart of a method for processing critical noise data provided in an embodiment of the present application is shown in FIG. Figure 4 As shown, a processing device applied to critical noise data, the method includes:

[0095] Step 401: Acquire critical noise point data of the surface of the object to be detected, wherein the critical noise point data includes a plurality of critical noise points;

[0096] Step 402: determining a weight sequence with a specific rule for each critical noise point, wherein the sequence is used to assign a weight to the critical noise point in the critical noise point data;

[0097] Step 403: Matching each of the critical noise points of the critical noise point data with a weight in a weight sequence having a specific regularity;

[0098] Exemplarily, the 9 critical noise points included in the critical noise point data are smoothed. For the first critical noise point y[0], the corresponding set of weight series is {17, 14, 11, 8, 5, 2, -1, -4, -7}; then the weight corresponding to y[0] is 17, the weight corresponding to y[1] is 14, the weight corresponding to y[2] is 11, the weight corresponding to y[3] is 8, the weight corresponding to y[4] is 5, the weight corresponding to y[5] is 2, the weight corresponding to y[6] is -1, the weight corresponding to y[7] is -4, and the weight corresponding to y[8] is -7.

[0099] Step 404: multiply each of the critical noise points by a corresponding weight value to obtain each weighted critical noise point;

[0100] Step 405: Accumulate the weighted critical noise points to obtain a first accumulated value;

[0101] Step 406: Accumulate the weights to obtain a second accumulated value;

[0102] Step 407: Divide the first accumulated value by the second accumulated value to obtain a weighted arithmetic mean value of the corresponding critical noise point, wherein all the calculated weighted arithmetic mean values ​​are determined as the smoothed critical noise point data;

[0103] Step 408: Determine the smoothed critical noise point data that is not within the first data range as defect data, wherein the first data range refers to a data range set according to a standard value and a design tolerance;

[0104] Step 409: removing the defect data from the smoothed critical noise data to obtain filtered critical noise data;

[0105] Step 410: Utilize the filtered critical noise point data to detect the surface of the object to be detected.

[0106] Exemplarily, at least two different non-average algorithms with weights are used to process 9 critical noise points. Each time a smoothed critical noise point is calculated, each of the 9 critical noise points corresponds to a set of weights, and this set of weights satisfies the characteristics of an arithmetic progression; for example, for the critical noise point y[0], the corresponding set of weights is {17, 14, 11, 8, 5, 2, -1, -4, -7}; for the critical noise point y[1], the corresponding set of weights is {56, 47, 38, 29, 20, 11, 2, -7, -16}; for the critical noise point y[2], the corresponding set of weights is {22, 19, 16, 13, 10, 7, 4, 1, -2}; for the critical noise point y[3], the corresponding set of weights is {32, 29, 26, 23, 20, 17 ,14,11,8}; for the critical noise point y[4], the corresponding set of weights is {1,1,1,1,1,1,1,1,1}; for the critical noise point y[5], the corresponding set of weights is {8,11,14,17,20,23,26,29,32}; for the critical noise point y[6], the corresponding set of weights is {-2,1,4,7,10,13,16,19,22}; for the critical noise point y[7], the corresponding set of weights is {-16,-7,2,11,20,29,38,47,56}; for the critical noise point y[8], the corresponding set of weights is {-7,-4,-1,2,5,8,11,14,17}.

[0107] The process of determining the smoothed critical noise data is as follows:

[0108] yy[0]=(17*y[0]+14*y[1]+11*y[2]+8*y[3]+5*y[4]+2*y[5]-y[6]-4*y[7]-7*y[8]) / 45;

[0109] yy[1]=(56*y[0]+47*y[1]+38*y[2]+29*y[3]+20*y[4]+11*y[5]+2*y[6]-7*y[7]-16*y[8]) / 180;

[0110] yy[2]=(22*y[0]+19*y[1]+16*y[2]+13*y[3]+10*y[4]+7*y[5]+4*y[6]+y[7]-2*y[8]) / 90;

[0111] yy[3]=(32*y[0]+29*y[1]+26*y[2]+23*y[3]+20*y[4]+17*y[5]+14*y[6]+11*y[7]+8*y[8]) / 180;

[0112] yy[4]=(y[0]+y[1]+y[2]+y[3]+y[4]+y[5]+y[6]+y[7]+y[8]) / 9;

[0113] yy[5]=(8*y[0]+11*y[1]+14*y[2]+17*y[3]+20*y[4]+23*y[5]+26*y[6]+29*y[7]+32*y[8]) / 180;

[0114] yy[6]=(-2*y[0]+y[1]+4*y[2]+7*y[3]+10*y[4]+13*y[5]+16*y[6]+19*y[7]+22*y[8]) / 90;

[0115] yy[7]=(-16*y[0]-7*y[1]+2*y[2]+11*y[3]+20*y[4]+29*y[5]+38*y[6]+47*y[7]+56*y[8]) / 180;

[0116] yy[8]=(-7*y[0]-4*y[1]-y[2]+2*y[3]+5*y[4]+8*y[5]+11*y[6]+14*y[7]+17*y[8]) / 45;

[0117] Among them, yy[0], yy[1], ..., yy[8] represent the 9 point cloud data after smoothing, and y[0], y[1], ..., y[8] represent the data values ​​of the 9 point cloud data before smoothing. By multiplying the 9 point cloud data by 9 different weights, 9 smoothed critical noise point data can be obtained.

[0118] In the embodiment of the present application, each critical noise point is matched with a weight in a weight sequence with a specific rule; each critical noise point is multiplied by the corresponding weight to obtain each weighted critical noise point; the weighted critical noise points are accumulated to obtain a first accumulated value; the weights are accumulated to obtain a second accumulated value; the first accumulated value is divided by the second accumulated value to obtain the weighted arithmetic mean of each weighted calculation. In this way, the critical noise data can be smoothed, the degree of deviation of the critical noise data from the real data is reduced, the polarity of the critical noise data is reduced, the critical noise data is more in line with the overall situation, more critical noise data is guaranteed to be within the standard data range, the amount of filtered critical noise data is reduced, the critical noise data corresponding to the real data is avoided from being misjudged and filtered, the real data is avoided from being missing, and the accuracy of defect analysis is improved.

[0119] The main fields involved in this embodiment are industrial intelligent detection and intelligent manufacturing, especially the analysis method of detecting plane defects and flaws of parts that appear in the industrial processing process.

[0120] In the actual use of existing plane defect detection, a 3D ranging sensor is usually used to obtain point cloud data, and the obtained point cloud data is quantitatively analyzed.

[0121] In the existing point cloud data quantitative analysis method, for obvious noise points in the data, the obvious noise data is removed. For data in the critical range of noise, some are removed and some are left. In the actual data analysis, the values ​​at the critical edge show their own characteristics; during the analysis process, the real critical noise data is often discarded, resulting in the inability to display the original physical characteristics of the defects / blemishes.

[0122] Existing methods include critical noise processing methods, which are theoretically feasible methods. However, in actual production, there are indeed very few problems caused by production, such as aging of production tools and temporary interruption of manufacturing due to failures. Therefore, when processing critical noise data, some real data will also be filtered, resulting in data loss. This has an irreversible impact on the subsequent result analysis.

[0123] This embodiment improves the existing method and improves the problem of over-killing critical noise data in the existing method. Through data screening, troubleshooting, and smoothing processing methods, the existing problems of data loss and small data volume in defect detection are effectively improved.

[0124] Figure 5A A flowchart of a method for processing critical noise data provided by an embodiment of the present application is shown below. Figure 5A As shown, the method includes:

[0125] Step 501, sensor data collection;

[0126] Step 502, the device filters the collected data;

[0127] Step 503, the device determines the validity of the filtered data;

[0128] After judging the validity of the data, there are three results: invalid data, valid data, and the validity of the data cannot be determined.

[0129] When data is judged to be invalid, it means that the data is obviously noisy data and needs to be screened out;

[0130] When the data is judged to be valid, it means that the data is valid data and not noise data, and step 505 is executed to directly output the value of the data;

[0131] When the data is determined to be incapable of determining validity, step 504 is executed;

[0132] Step 504, the device performs critical noise data processing on the data whose validity cannot be determined;

[0133] Here, the critical noise data processing is smoothing processing, and the smoothing processing is completed using a smoothing algorithm. The smoothing algorithm is: take the nth point of the sequence A, and take k values ​​before and after, and the value of n is not less than the value of k. The resulting sequence is {A(nk), A(n-k+1), ...A(n), ...A(n+k-1), A(n+k)}, and the total value is 2k+1 (the values ​​in brackets are data subscripts). Weights are added to the front of each value, and the weights are B(1), B(2), ..., B(2k+1). Therefore, after the current value is weighted and calculated, the resulting sequence is the smoothed value.

[0134] Step 505, the device outputs data;

[0135] Here, output data refers to outputting the smoothed critical noise data value.

[0136] Step 505: The device outputs a conclusion.

[0137] By comparing the output critical noise data value with the standard data range, it can be determined whether the data value is a defect. Here, the standard data range refers to the data range set according to the standard value and the design tolerance. When the data meets the standard data range, the data is not a defect and is retained; the data that is not within the range is a defect and needs to be filtered. After comparison, the operation performed on the data is output.

[0138] The defect detection method provided in this embodiment is mainly completed by a data collection module, an obvious noise data filtering module and a critical noise data algorithm processing module, wherein:

[0139] 1) Data collection module: two cameras at a certain distance from each other on the detection plane use a 3D ranging sensor to simultaneously acquire two images of the same scene. The corresponding pixels in the two images are found through a stereo matching algorithm. Then, the disparity information is calculated based on the triangulation principle and the pixel values ​​of the corresponding pixels in the two images. The disparity information is directly converted into planar point cloud data that can be used to characterize objects in the scene through the sensor. Figure 5B A schematic diagram of plane point cloud data of a product appearance surface provided in an embodiment of the present application is provided. The product appearance surface 510 is linearly scanned by a measuring instrument to obtain a point data set of the product appearance surface 510, form a dot matrix, and obtain plane point cloud data.

[0140] 2) Obvious noise data filtering module, including data deletion and data screening. Among them:

[0141] Data deletion: By comparing with the standard value and combining the design tolerance of the part itself, set the standard data range to remove obvious noise values. For example, if the standard value is 0.50 and the design tolerance is ±0.020, the standard data range is (0.48, 0.52). When the value is greater than 0.52 or less than 0.48, the data is in the non-standard data range and is considered as noise data. By setting the non-standard data range, filter directly and remove noise data. Here, the non-standard data range can be set according to different usage scenarios.

[0142] Data screening: Data with obvious noise can be filtered directly. By removing the obvious noise data, the remaining data is valid data with critical values. For example, if the standard value is 0.50, the design tolerance is ±0.020, and the tolerance requirement for the critical value is: ±0.015, that is, data with values ​​between 0.48 and 0.485, and between 0.515 and 0.52 are called critical value data.

[0143] 3) Critical noise data algorithm processing module, by smoothing the critical noise data, can make the real critical noise data within the data standard data range, reduce the misjudgment of critical noise data, and thus reduce the over-killing of critical noise data. Here, the smoothing process is completed using a smoothing algorithm.

[0144] Introduction to smoothing algorithm: Take the nth point of sequence A, and take k values ​​before and after, with n not less than k. The resulting sequence is A(nk), A(n-k+1),...A(n),...A(n+k-1),A(n+k), and the total value is 2k+1 (the values ​​in brackets are data subscripts). Add weights to the front of each value, with weights of B(1), B(2),..., B(2k+1). Therefore, after weighted calculation of the current value, the resulting sequence is C(n), and the current value is:

[0145] C(n)=[B(1)*A(nk)+B(2)*A(n-k+1)...+B(2k+1)*A(n+k)] / [B(1)+B(2)+...B(2k+1)]

[0146] Among them, C(n) is the weighted arithmetic mean; A(nk), A(n-k+1),...A(n),...A(n+k-1), A(n+k) are the values ​​of the nk to n+kth numbers in the sequence A; B(1), B(2),..., B(2k+1) are the weighted values ​​corresponding to these 2K+1 values.

[0147] If B(1)=B(2)=...=B(2K+1), it is called average weighting. For example, the commonly used 3-point and 5-point average weighting algorithms are:

[0148] 3-point average weighted: Cn = [A(n-1) + An + A(n+1)] / 3;

[0149] 5-point average weighted: Cn = [A(n-2)+A(n-1)+An+A(n+1)+A(n+2)] / 5;

[0150] If there are at least two different values ​​in the B sequence, we call it non-even weighting. For example, the commonly used 3-point and 5-point non-even weighting algorithms are:

[0151] 3-point non-average weighting: Cn = [2*A(n-1)+3*An+2*A(n+1)] / 7;

[0152] 5-point non-equal weighting: Cn = [2*A(n-2)+3*A(n-1)+4*An+3*A(n+1)+2*A(n+2)] / 14;

[0153] In some embodiments, a 9-point non-average weighted algorithm is used, and the calculation formula is as follows:

[0154] yy[0]=(17*y[0]+14*y[1]+11*y[2]+8*y[3]+5*y[4]+2*y[5]-y[6]-4*y[7]-7*y[8]) / 45;

[0155] yy[1]=(56*y[0]+47*y[1]+38*y[2]+29*y[3]+20*y[4]+11*y[5]+2*y[6]-7*y[7]-16*y[8]) / 180;

[0156] yy[2]=(22*y[0]+19*y[1]+16*y[2]+13*y[3]+10*y[4]+7*y[5]+4*y[6]+y[7]-2*y[8]) / 90;

[0157] yy[3]=(32*y[0]+29*y[1]+26*y[2]+23*y[3]+20*y[4]+17*y[5]+14*y[6]+11*y[7]+8*y[8]) / 180;

[0158] yy[4]=(y[0]+y[1]+y[2]+y[3]+y[4]+y[5]+y[6]+y[7]+y[8]) / 9;

[0159] yy[5]=(8*y[0]+11*y[1]+14*y[2]+17*y[3]+20*y[4]+23*y[5]+26*y[6]+29*y[7]+32*y[8]) / 180;

[0160] yy[6]=(-2*y[0]+y[1]+4*y[2]+7*y[3]+10*y[4]+13*y[5]+16*y[6]+19*y[7]+22*y[8]) / 90;

[0161] yy[7]=(-16*y[0]-7*y[1]+2*y[2]+11*y[3]+20*y[4]+29*y[5]+38*y[6]+47*y[7]+56*y[8]) / 180;

[0162] yy[8]=(-7*y[0]-4*y[1]-y[2]+2*y[3]+5*y[4]+8*y[5]+11*y[6]+14*y[7]+17*y[8]) / 45;

[0163] Among them, yy[0], yy[1], ..., yy[8] represent the 9 point cloud data after smoothing, y[0], y[1], ..., y[8] represent the data values ​​of these 9 point cloud data before smoothing, the weighted value sequence is an arithmetic sequence, and the required value maintains the characteristic of maximum weight.

[0164] Figure 5C A schematic diagram of the result of data smoothing for the critical noise data provided in the embodiment of the present application, wherein series 1 is the point cloud data that has not been smoothed; series 2 is the point cloud data that has been smoothed; the horizontal axis is the position of the nth point cloud data on the plane, wherein n is [0,8]; the vertical axis is the offset of the point cloud data from the standard value. Figure 5C As shown in the figure, through data smoothing, the original fluctuation becomes smaller, the peak-to-valley range difference is reduced, the broken line becomes smoother, the polarity is reduced, and the passing characteristic is reduced, that is, the curve is flatter and the data is more reliable. Among them, the passing characteristic reduction means that after setting the non-standard data range, the value of the critical noise data that meets the non-standard data range is reduced.

[0165] Based on the foregoing embodiments, an embodiment of the present application provides a device for processing critical noise data, which includes the units included and the modules included in the units, and can be implemented by a processor in a computer device; of course, it can also be implemented by a specific logic circuit; in the implementation process, the processor can be a central processing unit (CPU), a microprocessor (MPU), a digital signal processor (DSP) or a field programmable gate array (FPGA), etc.

[0166] Figure 6 A schematic diagram of the structure of a critical noise point data processing device provided in an embodiment of the present application is shown in FIG. Figure 6As shown, the device 600 includes an acquisition module 601, a processing module 602, a determination module 603, a filtering module 604 and a detection module 605, wherein:

[0167] An acquisition module 601 is used to acquire critical noise point data on the surface of the object to be detected;

[0168] A processing module 602 is used to smooth the critical noise point data to obtain smoothed critical noise point data;

[0169] A determination module 603, configured to determine the smoothed critical noise point data that is not within a first data range as defect data, wherein the first data range refers to a data range set according to a standard value and a design tolerance;

[0170] A filtering module 604 is used to remove the defect data from the smoothed critical noise data to obtain filtered critical noise data;

[0171] The detection module 605 is used to detect the surface of the object to be detected by using the filtered critical noise point data.

[0172] In some embodiments, the acquisition module 601 includes a detection submodule, a first determination submodule and a filtering submodule, wherein:

[0173] A detection submodule, used to obtain plane point cloud data of the surface of the object to be detected;

[0174] A first determination submodule is used to determine the plane point cloud data that is not within a second data range as noise data, wherein the second data range is determined by a standard value and a design tolerance; and the second data range is larger than the first data range;

[0175] The filtering submodule is used to filter out the noise point data from the plane point cloud data to obtain the critical noise point data.

[0176] In some embodiments, the filtering submodule includes a filtering unit and a first determining unit, wherein:

[0177] A filtering unit, used to filter out the noise data from the plane point cloud data to obtain first point cloud data;

[0178] The first determining unit is used to determine the first point cloud data that meets the third data range as critical noise point data, wherein the third data range is determined by the standard value, the design tolerance and the critical value tolerance.

[0179] In some embodiments, the detection submodule includes an acquisition unit, a matching unit, and a second determination unit, wherein:

[0180] An acquisition unit, used for acquiring at least two different images of the surface of the object to be detected in the same scene;

[0181] A matching unit, configured to match the at least two different images using a stereo matching algorithm to obtain corresponding pixel points in the at least two different images;

[0182] The second determining unit is used to determine the plane point cloud data according to the pixel value of the pixel point.

[0183] In some embodiments, the second determination unit is used to calculate the pixel value of the pixel point according to the triangulation principle to obtain disparity information corresponding to the pixel value; and convert the disparity information into plane point cloud data according to a specific corresponding relationship, wherein the specific corresponding relationship is used to characterize the corresponding relationship between the disparity information and the plane point cloud data.

[0184] In some embodiments, the critical noise point data includes a plurality of critical noise points, and the processing module 602 includes a second determination submodule and a third determination submodule, wherein:

[0185] A second determination submodule is used to determine a weight sequence with a specific rule for each critical noise point, wherein the sequence is used to assign a weight to the critical noise point in the critical noise point data;

[0186] The third determination submodule is used to determine the weighted arithmetic mean of the corresponding critical noise point according to the weight series and the critical noise point data to obtain the smoothed critical noise point data.

[0187] In some embodiments, the third determination submodule includes a corresponding unit, a multiplication unit, a first accumulation unit, a second accumulation unit and a division unit, wherein:

[0188] A corresponding unit, used for corresponding each critical noise point of the critical noise point data to a weight in a weight sequence having a specific regularity;

[0189] A multiplication unit, used for multiplying each of the critical noise points by a corresponding weight value to obtain each weighted critical noise point;

[0190] A first accumulation unit, used for accumulating the weighted critical noise points to obtain a first accumulated value;

[0191] A second accumulating unit, used for accumulating the weights to obtain a second accumulated value;

[0192] A division unit is used to divide the first accumulated value by the second accumulated value to obtain a weighted arithmetic mean value corresponding to a critical noise point.

[0193] The description of the above device embodiment is similar to the description of the above method embodiment, and has similar beneficial effects as the method embodiment. For technical details not disclosed in the device embodiment of the present application, please refer to the description of the method embodiment of the present application for understanding.

[0194] It should be noted that in the embodiment of the present application, if the above-mentioned method for processing critical noise data is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present application can be essentially or partly reflected in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device to execute all or part of the methods described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a U disk, a mobile hard disk, a read-only memory (ROM), a magnetic disk or an optical disk. In this way, the embodiment of the present application is not limited to any specific combination of hardware and software.

[0195] Correspondingly, an embodiment of the present application provides a critical noise data processing device, including a memory and a processor, wherein the memory stores a computer program that can be executed on the processor, and when the processor executes the program, the steps in the method provided in the above embodiment are implemented.

[0196] Correspondingly, an embodiment of the present application provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps in the method provided in the above embodiment are implemented.

[0197] It should be noted here that the description of the above storage medium and device embodiments is similar to the description of the above method embodiments, and has similar beneficial effects as the method embodiments. For technical details not disclosed in the storage medium and device embodiments of this application, please refer to the description of the method embodiments of this application for understanding.

[0198] It should be noted that Figure 7 A hardware entity schematic diagram of a critical noise data processing device provided in an embodiment of the present application, such as Figure 7 As shown, the hardware entity of the device 700 includes: a processor 701, a communication interface 702 and a memory 703, wherein

[0199] Processor 701 generally controls the overall operation of device 700 .

[0200] The communication interface 702 enables the device to communicate with other devices or servers through a network.

[0201] The memory 703 is configured to store instructions and applications executable by the processor 701, and can also cache data to be processed or processed by the processor 701 and each module in the device 700, which can be implemented through flash memory (FLASH) or random access memory (Random Access Memory, RAM).

[0202] It should be understood that "one embodiment" or "an embodiment" mentioned throughout the specification means that specific features, structures or characteristics related to the embodiment are included in at least one embodiment of the present application. Therefore, "in one embodiment" or "in an embodiment" appearing throughout the specification does not necessarily refer to the same embodiment. In addition, these specific features, structures or characteristics can be combined in one or more embodiments in any suitable manner. It should be understood that in various embodiments of the present application, the size of the sequence number of the above-mentioned processes does not mean the order of execution, and the execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application. The above-mentioned sequence numbers of the embodiments of the present application are only for description and do not represent the advantages and disadvantages of the embodiments.

[0203] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or device including the element.

[0204] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0205] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units; they may be located in one place or distributed on multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0206] In addition, all functional units in the embodiments of the present application may be integrated into one processing unit, or each unit may be a separate unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0207] A person skilled in the art can understand that all or part of the steps of implementing the above method embodiment can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiment; and the aforementioned storage medium includes: mobile storage devices, read-only memories (ROM), magnetic disks or optical disks, etc., various media that can store program codes.

[0208] Alternatively, if the above-mentioned integrated unit of the present application is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present application can essentially or in other words, the part that contributes to the relevant technology can be embodied in the form of a software product, which is stored in a storage medium and includes a number of instructions for a computer device to execute all or part of the methods described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, ROMs, magnetic disks or optical disks.

[0209] The above is only an implementation method of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for processing critical noise data, characterized in that: The method comprises: Acquire plane point cloud data of the surface of the object to be detected; Determine the plane point cloud data that is not within a second data range as noise data, wherein the second data range is determined by a standard value and a design tolerance; and the second data range is larger than the first data range; Filtering out the noise point data from the plane point cloud data to obtain critical noise point data; the critical noise point data includes a plurality of critical noise points; Determine a weight sequence with a specific rule for each critical noise point, wherein the sequence is used to assign weights to the critical noise points in the critical noise point data; determine a weighted arithmetic mean of the corresponding critical noise points according to the weight sequence and the critical noise point data, and obtain smoothed critical noise point data; Determine the smoothed critical noise point data that is not within a first data range as defect data, wherein the first data range refers to a data range set according to a standard value and a design tolerance; Removing the defect data from the smoothed critical noise point data to obtain filtered critical noise point data; Using the filtered critical noise point data to detect the surface of the object to be detected; Determining the weighted arithmetic mean of the corresponding critical noise point according to the weight series and the critical noise point data includes: Match each critical noise point of the critical noise point data with a weight in a weight sequence having a specific rule; multiply each critical noise point by the corresponding weight to obtain each weighted critical noise point; accumulate the weighted critical noise points to obtain a first accumulated value; accumulate the weights to obtain a second accumulated value; divide the first accumulated value by the second accumulated value to obtain a weighted arithmetic mean of the corresponding critical noise points.

2. The method according to claim 1, characterized in that The filtering out the noise data from the plane point cloud data to obtain the critical noise data includes: Filtering out the noise data from the plane point cloud data to obtain first point cloud data; The first point cloud data satisfying a third data range is determined as critical noise point data, wherein the third data range is determined by a standard value, a design tolerance, and a critical value tolerance.

3. The method according to claim 1, characterized in that The step of acquiring the plane point cloud data of the surface of the object to be detected comprises: Acquire at least two different images of the surface of the object to be detected in the same scene; Matching the at least two different images using a stereo matching algorithm to obtain corresponding pixel points in the at least two different images; Plane point cloud data is determined according to the pixel value of the pixel point.

4. The method according to claim 3, characterized in that The step of determining the plane point cloud data according to the pixel value of the pixel point comprises: Calculating the pixel value of the pixel point according to the triangulation principle to obtain the disparity information corresponding to the pixel value; The disparity information is converted into plane point cloud data according to a specific corresponding relationship, wherein the specific corresponding relationship is used to characterize the corresponding relationship between the disparity information and the plane point cloud data.

5. A device for processing critical noise data, characterized in that: The device comprises: An acquisition module is used to acquire plane point cloud data of the surface of the object to be detected; determine the plane point cloud data that is not within a second data range as noise data, wherein the second data range is determined by a standard value and a design tolerance; the second data range is larger than the first data range; filter out the noise data from the plane point cloud data to obtain critical noise data; the critical noise data includes a plurality of critical noise points; A processing module, used to determine a weight sequence with a specific rule for each critical noise point, wherein the sequence is used to assign weights to the critical noise points in the critical noise point data; and to determine a weighted arithmetic mean of the corresponding critical noise points according to the weight sequence and the critical noise point data to obtain smoothed critical noise point data; A determination module, configured to determine the smoothed critical noise point data that is not within a first data range as defective data, wherein the first data range refers to a data range set according to a standard value and a design tolerance; A filtering module, used to remove the defect data from the smoothed critical noise data to obtain filtered critical noise data; A detection module, used to detect the surface of the object to be detected using the filtered critical noise point data; The processing module is also used to correspond each critical noise point of the critical noise point data to a weight in a weight series with a specific rule; multiply each critical noise point by the corresponding weight to obtain each weighted critical noise point; accumulate the weighted critical noise points to obtain a first accumulated value; accumulate the weights to obtain a second accumulated value; divide the first accumulated value by the second accumulated value to obtain a weighted arithmetic mean of the corresponding critical noise points.

6. A critical noise data processing device, comprising a memory and a processor, wherein the memory stores a computer program that can be run on the processor, characterized in that: When the processor executes the program, the steps in the method according to any one of claims 1 to 4 are implemented.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps in the method according to any one of claims 1 to 4 are implemented.

Images