Method for judging accuracy of pavement disease recognition result and related equipment

By calculating the confidence interval and mean error of the repair and crack identification results, and combining them with the pavement damage index, the accuracy of the pavement defect identification results is automatically determined. This solves the problem of low defect identification accuracy in existing technologies, improves identification efficiency, and reduces the cost of manual review.

CN119004049APending Publication Date: 2024-11-22ROADMAINT CO LTD
View PDF 3 Cites 0 Cited by

Patent Information

Application Number
CN202411482362.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-10-23
Publication Date
2024-11-22

AI Technical Summary

Technical Problem

In existing technologies, the accuracy of pavement defect identification results is difficult to reach 100%, which leads to the need for manual verification in actual engineering applications, increasing costs.

Method used

By calculating the confidence intervals of the repair identification results and the baseline results, the mean area error of the crack identification results, and combining them with the pavement damage index, the boundary values ​​of the acceptance and rejection regions are determined, and the accuracy of the pavement distress identification results is automatically judged.

Benefits of technology

It has improved the accuracy of road surface recognition algorithm results by automatically judging them, improved the efficiency of defect identification, and reduced the cost of manual review.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119004049A_ABST
    Figure CN119004049A_ABST
Patent Text Reader

Abstract

The invention provides a method for judging the accuracy of a pavement disease recognition result and related equipment. The method comprises the following steps: calculating a confidence interval of a repair area according to an obtained repair identification result and a repair reference result; based on the obtained crack identification result and the crack reference result, calculating a crack area error mean value, and taking the crack area error mean value as a test statistic; obtaining a pavement damage condition index, and obtaining a first acceptable range of a total disease area error mean value according to the pavement damage condition index; determining an acceptance domain threshold value and a rejection domain threshold value based on the confidence interval and the first acceptable range; based on the acceptance domain threshold value and the rejection domain threshold value, obtaining a rejection domain; and in response to the test statistics falling into the rejection domain, determining that the pavement disease recognition result corresponding to the test statistics is inaccurate. Through the method provided by the invention, the accuracy of the pavement algorithm recognition result can be automatically judged, the pavement disease recognition efficiency is improved, and the manual rechecking cost is reduced.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a method and related equipment for determining the accuracy of road surface disease identification results. Background Art

[0002] With the development of high-performance computers and artificial intelligence technology, the degree of automation in pavement disease identification is becoming increasingly higher. The Pavement Condition Index (PCI) is an indicator that characterizes the pavement damage condition, which is calculated based on the pavement disease identification results. Therefore, the accuracy of pavement disease identification directly affects the accuracy of pavement PCI.

[0003] At present, it is difficult to achieve 100% accuracy in disease identification. In actual engineering applications, it is necessary to manually review the disease identification results to determine whether the disease identification results meet the requirements. The cost of manually reviewing a large number of disease identification results is high.

[0004] Therefore, how to automatically determine the accuracy of pavement algorithm recognition results in order to improve the efficiency of pavement disease recognition and reduce the labor cost of reviewing pavement disease recognition results is an urgent problem to be solved. Summary of the invention

[0005] In view of this, the purpose of this application is to propose a method and related equipment for determining the accuracy of road surface disease identification results, so as to solve or partially solve the above-mentioned problems.

[0006] Based on the above purpose, in a first aspect, the present application provides a method for determining the accuracy of a pavement defect recognition result, comprising: Acquire a pavement disease identification result and a pavement disease benchmark result, wherein the pavement disease identification result includes at least one of a repair identification result or a crack identification result, and the benchmark result includes at least one of a repair benchmark result or a crack benchmark result; Calculating a confidence interval of a repair area according to the repair identification result and the repair benchmark result; Based on the crack identification result and the crack benchmark result, calculating a mean value of crack area errors, and using the mean value of crack area errors as a test statistic; Obtaining a pavement damage condition index, and obtaining a first acceptable range of a mean error of an overall diseased area according to the pavement damage condition index; Determine an acceptance domain limit value and a rejection domain limit value based on the confidence interval and the first acceptable range; Based on the acceptance domain boundary value and the rejection domain boundary value, a rejection domain is obtained; In response to the test statistic falling into the rejection region, it is determined that a pavement defect identification result corresponding to the test statistic is inaccurate.

[0007] In a second aspect of the present application, a device for determining the accuracy of a road surface disease recognition result is provided, comprising: an acquisition module configured to acquire a pavement disease identification result and a pavement disease benchmark result, wherein the pavement disease identification result includes at least one of a repair identification result or a crack identification result, and the benchmark result includes at least one of a repair benchmark result or a crack benchmark result; A first calculation module is configured to calculate a confidence interval of a repair area according to the repair identification result and the repair reference result; A second calculation module is configured to calculate a mean value of crack area errors based on the crack identification result and the crack reference result, and use the mean value of crack area errors as a test statistic; A third calculation module is configured to obtain a pavement damage condition index, and obtain a first acceptable range of a mean error of an overall diseased area according to the pavement damage condition index; a fourth calculation module, configured to determine an acceptance domain boundary value and a rejection domain boundary value based on the confidence interval and the first acceptable range; A fifth calculation module is configured to obtain a rejection domain based on the acceptance domain boundary value and the rejection domain boundary value; The discrimination module is configured to determine that a pavement damage identification result corresponding to the test statistic is inaccurate in response to the test statistic falling into the rejection domain.

[0008] In a third aspect of the present application, an electronic device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the method described in the first aspect is implemented.

[0009] According to a fourth aspect of the present application, a non-transitory computer-readable storage medium is provided, wherein the non-transitory computer-readable storage medium stores computer instructions, wherein the computer instructions are used to enable a computer to execute the method described in the first aspect.

[0010] From the above, it can be seen that the present application provides a method and related equipment for determining the accuracy of pavement disease identification results. The method includes: calculating the confidence interval of the repair area based on the obtained repair identification results and repair benchmark results; calculating the mean crack area error based on the obtained crack identification results and crack benchmark results, and using the mean crack area error as the test statistic; obtaining the pavement damage condition index, and obtaining the first acceptable range of the overall disease area error mean based on the pavement damage condition index; determining the acceptance domain boundary value and the rejection domain boundary value based on the confidence interval and the first acceptable range; obtaining the rejection domain based on the acceptance domain boundary value and the rejection domain boundary value; in response to the test statistic falling into the rejection domain, determining that the pavement disease identification result corresponding to the test statistic is inaccurate. Through the method of the present application, the accuracy of the pavement algorithm identification results can be automatically determined, the efficiency of pavement disease identification can be improved, and the cost of manual review can be reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] In order to more clearly illustrate the technical solutions in the present application or related technologies, the drawings required for use in the embodiments or related technical descriptions are briefly introduced below. Obviously, the drawings described below are only embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0012] Figure 1 A schematic diagram of an exemplary system according to an embodiment of the present application is shown.

[0013] Figure 2 A schematic diagram showing exemplary pavement damage identification results according to an embodiment of the present application is shown.

[0014] Figure 3 A flow chart of an exemplary method for determining the accuracy of road surface defect identification results according to an embodiment of the present application is shown.

[0015] Figure 4 A schematic diagram of an exemplary device for determining the accuracy of road surface disease identification results according to an embodiment of the present application is shown.

[0016] Figure 5 A schematic diagram of an exemplary electronic device according to an embodiment of the present application is shown. DETAILED DESCRIPTION

[0017] In order to make the objectives, technical solutions and advantages of the present application more clearly understood, the present application is further described in detail below in combination with specific embodiments and with reference to the accompanying drawings.

[0018] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the embodiments of the present application should be the usual meanings understood by people with ordinary skills in the field to which the present application belongs. The "first", "second" and similar words used in the embodiments of the present application do not represent any order, quantity or importance, but are only used to distinguish different components. "Including" or "comprising" and similar words mean that the elements or objects appearing in front of the word cover the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0019] As mentioned above, at present, the accuracy of automatic disease identification is difficult to reach 100%. In actual engineering applications, it is necessary to manually review the automatic disease identification results to determine whether the automatic disease identification results meet the requirements. The cost of manually reviewing a large number of automatic disease identification results is high.

[0020] Therefore, how to automatically determine the accuracy of pavement algorithm recognition results in order to improve the efficiency of pavement disease recognition and reduce the labor cost of reviewing pavement disease recognition results is an urgent problem to be solved.

[0021] In order to at least solve the above problems, the present application provides a method and related equipment for determining the accuracy of pavement disease identification results. The method includes: calculating the confidence interval of the repair area based on the obtained repair identification results and repair benchmark results; calculating the mean of the crack area error based on the obtained crack identification results and crack benchmark results, and using the mean of the crack area error as the test statistic; obtaining the pavement damage condition index, and obtaining the first acceptable range of the mean of the overall disease area error based on the pavement damage condition index; determining the acceptance domain boundary value and the rejection domain boundary value based on the confidence interval and the first acceptable range; obtaining the rejection domain based on the acceptance domain boundary value and the rejection domain boundary value; in response to the test statistic falling into the rejection domain, determining that the pavement disease identification result corresponding to the test statistic is inaccurate. Through the method of the present application, the accuracy of the pavement algorithm identification results can be automatically determined, the efficiency of pavement disease identification can be improved, and the cost of manual review can be reduced.

[0022] Figure 1 A schematic diagram of an exemplary system 100 according to an embodiment of the present application is shown.

[0023] like Figure 1As shown, the system 100 may include a server 102 and terminal devices 104, 106, and 108, wherein the terminal devices 104, 106, and 108 are connected to the server 102 via a network, such as a wired or wireless network connection. In the embodiment of the present application, the apparatus for processing abnormal information of an application may be integrated in the terminal devices 104, 106, and 108.

[0024] The server side 102 may be a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN (Content Delivery Network), and big data and artificial intelligence platforms. The server side 102 may provide background services for the terminal device 104, the terminal device 106, and the terminal device 108. The operating system running on the server side 102 may include, but is not limited to, Android system, IOS system, Linux system, Windows system, etc.

[0025] The terminal device 104, the terminal device 106, and the terminal device 108 may include physical devices such as smart phones, desktop computers, tablet computers, laptop computers, smart speakers, vehicle-mounted terminals, digital assistants, augmented reality (AR) or virtual reality (VR) devices, smart wearable devices, etc., and may also include software running in the physical devices, such as applications, etc. In the embodiments of the present application, the operating systems running on the above-mentioned terminal devices 104, the terminal devices 106, and the terminal devices 108 may include, but are not limited to, Android systems, IOS systems, Linux systems, Windows systems, etc.

[0026] Various applications (Application, APP, Chinese abbreviation: application) can be installed on the terminal device 104, the terminal device 106, and the terminal device 108, such as image processing applications, video conferencing applications, reading applications, video applications, social applications, payment applications, web browsers, and instant messaging tools.

[0027] The above-mentioned terminal device 104, terminal device 106, terminal device 108 and server 102 can be used to implement the method for determining the accuracy of pavement disease identification results provided in the present application. For example, a user can send a pavement disease identification result and a pavement disease baseline result to the server 102 through the terminal device 104, terminal device 106 or terminal device 108, and the server 102 returns the accuracy determination result of the pavement disease identification result and displays it to the user through the terminal device 104, terminal device 106 or terminal device 108.

[0028] Figure 2 A schematic diagram of an exemplary pavement disease identification result 200 according to an embodiment of the present application is shown.

[0029] First, an automatic recognition algorithm is used to identify cracks and repairs in each road surface image in all road surface images to obtain road surface disease recognition results. The specific contents are as follows: The road surface image can be collected by a multifunctional road condition collection system. The resolution of the collected road surface image is set by the collection device, and the road surface image can be pre-processed in advance. The crack and repair information in the road surface image can be marked by an algorithm, and the marking method can be pixel or grid.

[0030] For example, Figure 2 As shown in the figure, the resolution of the original collected road image is 3712×1941 pixels, then pixels are filled on the right and bottom of the road image to fill the image with 3800×2000 pixels, where the grayscale value of the filled pixels is 255; the pixels of the marked square sub-image block are 100×100, Figure 2 The red and blue colors represent the pavement damage identification results of cracks and repairs, respectively.

[0031] It should be noted that the image recognition device can be a computer device in various forms, running an image recognition algorithm, such as image preprocessing, image binarization, crack location, etc., to identify road surface repairs, cracks and other defects, and save the recognition results. It is understood that the road surface defects identified by the image recognition device may also include block cracks, cracks, potholes, looseness, etc., which are not limited in this application.

[0032] According to a predetermined sampling ratio, a corresponding number of road surface images are extracted from all the road surface images in each assessment unit, and the automatic recognition results of cracks and repairs of each road surface image extracted from each assessment unit are manually reviewed to obtain a benchmark result for extracting cracks and repairs of each road surface image in each assessment unit; the specific contents are as follows: according to a fixed ratio, road surface images within the scope of each assessment unit are extracted, and on the basis of the existing automatic recognition results, the missed recognition or misrecognition in the images is modified through a manual review method, and this is used as a benchmark result for extracting cracks and repairs in road surface images.

[0033] The sampling method may be random sampling, equally spaced sampling, stratified plus random sampling, segmented sampling, etc., and this application does not limit this.

[0034] For example, we can take 1 km as the length of the assessment unit, and the interval between each road surface image is 2m, so each assessment unit has 500 images. The corresponding number of road surface images are randomly selected at a ratio of 10%. The algorithm and manual recognition results of 50 images in a certain assessment unit (the area of ​​each grid is the same, in units of grid numbers) are shown in Table 1.

[0035] Table 1: Road surface image sampling results.

[0036]

[0037] When judging the accuracy of the road surface disease recognition results, in some embodiments, the mean error and error variance of the repair area of ​​the extracted images can be calculated based on the benchmark recognition results after automatic and manual review of the road surface images extracted by each evaluation unit, and the parameter estimation is performed based on the repair area error of the extracted images to obtain the confidence interval of the mean error of the repair area under a certain confidence level. The specific contents are as follows: The mean error of the sample repair area is:

[0038] in, For the The automatic recognition result of the patching area of ​​the picture. For the The baseline results of the inpainting area of ​​the image, The number of images extracted for the rating unit.

[0039] The sampling evaluation error of non-repeated sampling is:

[0040] in, is the sample patching area error variance, is the total number of images in the evaluation unit. When the number of sample sets is greater than 30, according to the law of large numbers and the central limit theorem, the sampling distribution of the sample mean is approximately normal distribution, and the sampling limit error , the confidence interval is: .in, is the probability, is the minimum value of the confidence interval, is the maximum value of the confidence interval.

[0041] For example, according to the evaluation unit extraction image algorithm and manual results in Table 1, the repair area error in grid units is calculated as is 0.98, and the variance The sampling error is 30.84. is 0.75. If the significance level is 0.05, then the probability =1.96, and the confidence interval is: [-0.48,2.44].

[0042] After the confidence interval is calculated, in some embodiments, the mean value of the crack area error of each evaluation unit can be selected as the test statistic, and the acceptance domain and rejection domain according to the PCI error limit are determined. The test statistic is determined based on the sample data whether it falls into the above-mentioned rejection domain. If it does, the original hypothesis is rejected, and it is considered that the PCI value error of this evaluation unit is unacceptable; otherwise, it is considered that the PCI value of this evaluation unit is within an acceptable range. It can be understood that when the test statistic falls into the rejection domain, it is recorded as event A; when it falls into the acceptance domain, it is recorded as event B. Event A and event B are mutually opposed. Falling into the rejection domain is equivalent to not falling into the acceptance domain; conversely, not falling into the rejection domain is equivalent to falling into the acceptance domain.

[0043] Since the conversion coefficient of the repair area is relatively small compared to the crack, the mean error of the crack area with a larger conversion coefficient can be used as a test statistic to facilitate the identification of pavement disease results. The specific contents are as follows: In PCI calculation, the conversion factor of the repair area is 0.1, which is relatively small compared to the crack. First, the repair area error is estimated based on the pictures extracted from each assessment unit, and the hypothesis test of the mean error of the overall disease area is converted into a hypothesis test of the mean error of the crack area, that is:

[0044] in, is the mean value of the overall crack area error taking into account the conversion factor, To determine the acceptance domain based on PCI and the Distress Rate (DR) and the PCI error limit, is the corresponding rejection region, is the acceptance domain limit value, is the rejection region limit value. Based on the algorithm recognition, the PCI value of the assessment unit can be obtained. According to this PCI value, the functional relationship between PCI and damage rate, and the pre-set PCI tolerance, the acceptable range of DR can be calculated: ,in, is the minimum damage rate, is the maximum damage rate. The acceptable range of DR can be converted into the acceptable range of the mean error of the overall diseased area taking into account the conversion factor, that is, Acceptable range: ,in, is the minimum mean error of the total diseased area, is the maximum overall disease area error mean, is the conversion factor, This is the automatic identification result of the overall disease area. This is the baseline result for the overall diseased area.

[0045] Understandably, and Represent the dividing points of the acceptance domain and the rejection domain, respectively, which determine the critical value of whether to reject the null hypothesis. Specifically, It is the lower boundary point where the statistic enters the rejection region. That is, when the statistic value is lower than or equal to this value, the null hypothesis is rejected. This value is the lower boundary of the acceptance region. It is the upper boundary point where the statistic enters the rejection region. That is, when the statistic value is higher than or equal to this value, the null hypothesis should also be rejected. This value is the upper boundary of the acceptance region.

[0046] It should be noted that PCI and DR can be converted using the following formula:

[0047]

[0048] Among them, for the asphalt pavement coefficient Use 15 for the cement concrete pavement coefficient Use 10.66 for the asphalt pavement index Use 0.412 for cement concrete pavement index Using 0.461, For the The cumulative area of ​​road damage in each picture, For road surface inspection or survey area, is the upper limit value.

[0049] The cumulative area of ​​road damage can be calculated using the following formula:

[0050] in, For the The number of grids with damaged road surface in the image, is the area conversion factor.

[0051] Based on confidence interval and acceptable range , the acceptance domain boundary value can be determined and the rejection region limit value .

[0052] For example, taking the assessment unit in Table 1 as an example, the PCI obtained based on the algorithm recognition result is 73.71, and the preset PCI allowable error is 3, then the PCI allowable error range is [70.71, 76.71], and the algorithm recognition result of the assessment unit can pass; according to the functional relationship between PCI and DR, the corresponding overall disease error mean range is calculated inversely is: [-7.87, 6.68]; further we get is -7.82, It is 6.44.

[0053] Specifically, the acceptance region boundary value and the rejection region boundary value can be obtained by the following method.

[0054] In some embodiments, the observed value in the hypothesis test is the crack area error, that is:

[0055] in, For the The crack area of ​​the image is automatically identified. For the Benchmark results for crack area in Figure 1.

[0056] The parameter space corresponding to the original hypothesis is:

[0057] Need help Compare with other known conditions , By estimating, we can get: =

[0058] Among them, the conversion factor , Conversion factor , For the The crack area of ​​the image is automatically identified. For the Benchmark results for crack area in Figure 1.

[0059] according to , , combined with the confidence interval of the mean error of the repair area under a certain confidence level, we can get: , .

[0060] Based on the acceptance region limit value and the rejection region limit value, the rejection region can be obtained:

[0061] in, is the acceptance domain limit value, is the rejection region limit value, is the mean error of the crack area, represents a t distribution with n-1 degrees of freedom ( is the critical value of the t-distribution, or percentile), S is the sample variance, n is the sample size, The probability of making a Type I error.

[0062] Determine whether the test statistic falls into the above rejection region based on the sample crack data. If so, it is considered that the PCI value error of this assessment unit is unacceptable; otherwise, it is considered that the PCI value of this assessment unit is within an acceptable range.

[0063] Specifically, the rejection region can be obtained by the following method.

[0064] The hypothesis test of the population by the sample is always completed by a statistic, which is called the test statistic. In some embodiments, the mean error of the sample crack area can be selected. As a test statistic, its rejection region is in the following form:

[0065] in, and represent the cutoff points of the acceptance domain and the rejection domain, and and The same in a conceptual sense. and is the dividing point corresponding to the mean error of the sample crack area, and is the dividing point corresponding to the mean value of the overall crack area error, represents the mean error.

[0066] In hypothesis testing, if ( is the mean error of the overall crack area, is the parameter space corresponding to the null hypothesis), the probability of making a first-type error is α(θ)<α (α is the significance level), then the null hypothesis is accepted at this significance level, that is, the PCI value is considered to be within the allowable error range. The probability of making a first-type error is:

[0067] in, is a function symbol, for example, It can be a probability function.

[0068] Right now:

[0069] random variable It obeys the t distribution with n-1 degrees of freedom (where S is the sample variance and n is the sample size), that is, , and the rejection region is derived. It is obtained by replacing the independent variable and is the dividing point between the acceptance domain and the rejection domain after changing to the corresponding statistic.

[0070] According to the evaluation unit extraction image algorithm and manual results in Table 1, the mean error of the sample crack area in grid units is calculated. is 2.10, and the variance is 30.84, and the significance level α is 0.1, based on which the rejection region is obtained Obviously, 2.1 does not fall into the rejection region, and the original hypothesis is considered reasonable, that is, the PCI error of this kilometer is within the acceptable range.

[0071] In order to verify the accuracy of the determination method, all pictures of the kilometer were manually reviewed and the PCI value was 76.22. The error between the algorithm PCI value and the manual PCI was -2.5, which was within the allowable error range, indicating the rationality of the determination method proposed in this patent.

[0072] The final output of the pavement disease identification result is the PCI calculated by the assessment unit. Based on the manual review of the assessment unit image sampling and PCI error control, the mathematical and statistical methods are used to determine whether the automatic identification of the assessment unit meets the requirements, which will greatly save the cost of manual review.

[0073] Figure 3 FIG. 3 is a flow chart of an exemplary method 300 for determining the accuracy of a road surface disease identification result according to an embodiment of the present application. Figure 3 As shown, method 300 may include the following steps.

[0074] In step 302, a pavement disease identification result and a pavement disease benchmark result are obtained, wherein the pavement disease identification result includes at least one of a repair identification result or a crack identification result, and the benchmark result includes at least one of a repair benchmark result or a crack benchmark result.

[0075] In step 304, a confidence interval of the repair area is calculated according to the repair identification result and the repair reference result.

[0076] In some embodiments, the calculation of the confidence interval of the patch area based on the patch identification result and the patch benchmark result further includes: calculating the mean patch area error and the variance of the patch area error based on the patch identification result and the patch benchmark result; and calculating the confidence interval based on the mean patch area error and the variance of the patch area error.

[0077] In some embodiments, the mean value of the patching area error is expressed as:

[0078] in, is the patch recognition result of the i-th image, is the inpainting benchmark result of the i-th image, and n is the number of images.

[0079] In some embodiments, the calculating the confidence interval based on the mean of the patch area error and the variance of the patch area error further includes: obtaining an evaluation error according to the variance of the patch area error; and calculating the confidence interval based on the evaluation error and the mean of the patch area error.

[0080] In step 306, based on the crack identification result and the crack reference result, a mean value of the crack area error is calculated, and the mean value of the crack area error is used as a test statistic.

[0081] In step 308, a pavement damage condition index is obtained, and a first acceptable range of the mean error of the overall diseased area is obtained according to the pavement damage condition index.

[0082] In some embodiments, the first acceptable range of the mean error of the overall diseased area obtained according to the pavement damage condition index further includes: obtaining a second acceptable range of pavement damage rate according to the pavement damage condition index; and converting the second acceptable range into the first acceptable range.

[0083] In step 310, based on the confidence interval and the first acceptable range, an acceptance region limit value and a rejection region limit value are determined.

[0084] In some embodiments, The acceptance domain limit value is expressed as: , The rejection region boundary value is expressed as: , in, is the first acceptable range, is the minimum mean error of the total diseased area, is the maximum overall disease area error mean, is the confidence interval, is the minimum value of the confidence interval, is the maximum value of the confidence interval.

[0085] In step 312, a rejection region is obtained based on the acceptance region boundary value and the rejection region boundary value.

[0086] In some embodiments, the rejection region is represented by:

[0087] in, is the acceptance domain limit value, is the rejection region limit value, is the mean error of the crack area, represents the t distribution with n-1 degrees of freedom, S is the sample variance, n is the sample size, The probability of making a Type I error.

[0088] In step 314 , in response to the test statistic falling into the rejection region, it is determined that the pavement defect identification result corresponding to the test statistic is inaccurate.

[0089] The present application provides a method and related equipment for determining the accuracy of pavement disease identification results. The method includes: calculating the confidence interval of the repair area based on the obtained repair identification results and repair benchmark results; calculating the mean of the crack area error based on the obtained crack identification results and crack benchmark results, and using the mean of the crack area error as the test statistic; obtaining the pavement damage condition index, and obtaining the first acceptable range of the mean of the overall disease area error based on the pavement damage condition index; determining the acceptance domain boundary value and the rejection domain boundary value based on the confidence interval and the first acceptable range; obtaining the rejection domain based on the acceptance domain boundary value and the rejection domain boundary value; and determining that the pavement disease identification result corresponding to the test statistic is inaccurate in response to the test statistic falling into the rejection domain. Through the method of the present application, the accuracy of the pavement algorithm identification results can be automatically determined, the efficiency of pavement disease identification can be improved, and the cost of manual review can be reduced.

[0090] It should be noted that the method of the embodiment of the present application can be performed by a single device, such as a computer or server. The method of this embodiment can also be applied to a distributed scenario and completed by multiple devices cooperating with each other. In the case of such a distributed scenario, one of the multiple devices can only perform one or more steps in the method of the embodiment of the present application, and the multiple devices will interact with each other to complete the described method.

[0091] It should be noted that the above describes some embodiments of the present application. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the above embodiments and still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0092] Based on the same technical concept, corresponding to any of the above-mentioned embodiment methods, the present application also provides a device for determining the accuracy of road hazard identification results.

[0093] refer to Figure 4 The device for determining the accuracy of the road surface disease identification result comprises: The acquisition module 401 is configured to acquire a pavement disease identification result and a pavement disease benchmark result, wherein the pavement disease identification result includes at least one of a repair identification result or a crack identification result, and the benchmark result includes at least one of a repair benchmark result or a crack benchmark result.

[0094] The first calculation module 402 is configured to calculate a confidence interval of a repair area according to the repair identification result and the repair reference result.

[0095] The first calculation module 402 is further configured to calculate a mean value of a patch area error and a variance of a patch area error according to the patch identification result and the patch reference result; and calculate the confidence interval based on the mean value of the patch area error and the variance of the patch area error.

[0096] Wherein, the mean value of the repair area error is expressed as:

[0097] in, is the patch recognition result of the i-th image, is the inpainting benchmark result of the i-th image, and n is the number of images.

[0098] The first calculation module 402 is further configured to obtain an evaluation error according to the repair area error variance; and calculate the confidence interval based on the evaluation error and the repair area error mean.

[0099] The second calculation module 403 is configured to calculate a mean value of crack area errors based on the crack identification result and the crack reference result, and use the mean value of crack area errors as a test statistic.

[0100] The third calculation module 404 is configured to obtain a pavement damage condition index, and obtain a first acceptable range of a mean error of the overall diseased area according to the pavement damage condition index.

[0101] The third calculation module 404 is further configured to obtain a second acceptable range of the road surface damage rate according to the road surface damage condition index; and convert the second acceptable range into the first acceptable range.

[0102] The fourth calculation module 405 is configured to determine an acceptance region boundary value and a rejection region boundary value based on the confidence interval and the first acceptable range.

[0103] Wherein, the acceptance domain limit value is expressed as: , The rejection region boundary value is expressed as: , in, is the first acceptable range, is the confidence interval.

[0104] The fifth calculation module 406 is configured to obtain a rejection domain based on the acceptance domain boundary value and the rejection domain boundary value.

[0105] Wherein, the rejection region is expressed as:

[0106] in, is the acceptance domain limit value, is the rejection region limit value, is the mean error of the crack area, represents the t distribution with n-1 degrees of freedom, S is the sample variance, n is the sample size, The probability of making a Type I error.

[0107] The determination module 407 is configured to determine that the pavement damage identification result corresponding to the test statistic is inaccurate in response to the test statistic falling into the rejection region.

[0108] For the convenience of description, the above device is described in terms of functions divided into various modules. Of course, when implementing the present application, the functions of each module can be implemented in the same or multiple software and / or hardware.

[0109] The device of the above embodiment is used to implement the corresponding method 300 in any of the above embodiments, and has the beneficial effects of the corresponding method embodiment, which will not be described in detail here.

[0110] Based on the same technical concept, corresponding to any of the above-mentioned embodiments, the present application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the method 300 described in any of the above embodiments when executing the program.

[0111] Figure 5 A schematic diagram of an exemplary electronic device according to an embodiment of the present application is shown, and the device may include: a processor 1010, a memory 1020, an input / output interface 1030, a communication interface 1040, and a bus 1050. The processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040 are connected to each other in communication within the device through the bus 1050.

[0112] The processor 1010 can be implemented by a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this specification.

[0113] The memory 1020 may be implemented in the form of ROM (Read Only Memory), RAM (Random Access Memory), static storage device, dynamic storage device, etc. The memory 1020 may store an operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented by software or firmware, the relevant program codes are stored in the memory 1020 and are called and executed by the processor 1010.

[0114] The input / output interface 1030 is used to connect the input / output module to realize information input and output. The input / output module can be configured in the device as a component (not shown in the figure), or it can be externally connected to the device to provide corresponding functions. The input device may include a keyboard, a mouse, a touch screen, a microphone, various sensors, etc., and the output device may include a display, a speaker, a vibrator, an indicator light, etc.

[0115] The communication interface 1040 is used to connect a communication module (not shown) to realize communication interaction between the device and other devices. The communication module can realize communication through wired means (such as USB, network cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.).

[0116] The bus 1050 includes a path that transmits information between the various components of the device (eg, the processor 1010 , the memory 1020 , the input / output interface 1030 , and the communication interface 1040 ).

[0117] It should be noted that, although the above device only shows the processor 1010, the memory 1020, the input / output interface 1030, the communication interface 1040 and the bus 1050, in the specific implementation process, the device may also include other components necessary for normal operation. In addition, it can be understood by those skilled in the art that the above device may also only include the components necessary for implementing the embodiments of the present specification, and does not necessarily include all the components shown in the figure.

[0118] The electronic device of the above embodiment is used to implement the corresponding method 300 in any of the above embodiments, and has the beneficial effects of the corresponding method embodiment, which will not be described in detail here.

[0119] Based on the same technical concept, corresponding to any of the above-mentioned embodiments, the present application also provides a non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute method 300 described in any of the above embodiments.

[0120] The computer-readable medium of this embodiment includes permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, read-only compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device.

[0121] The computer instructions stored in the storage medium of the above embodiment are used to enable the computer to execute the method 300 described in any of the above embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0122] A person skilled in the art should understand that the discussion of any of the above embodiments is merely illustrative and is not intended to imply that the scope of the present application (including the claims) is limited to these examples. In line with the concept of the present application, the technical features in the above embodiments or different embodiments may be combined, the steps may be implemented in any order, and there are many other variations of the different aspects of the embodiments of the present application as described above, which are not provided in detail for the sake of simplicity.

[0123] In addition, to simplify the description and discussion, and in order not to make the embodiments of the present application difficult to understand, the known power / ground connections to the integrated circuit (IC) chip and other components may or may not be shown in the provided drawings. In addition, the device may be shown in the form of a block diagram to avoid making the embodiments of the present application difficult to understand, and this also takes into account the fact that the details of the implementation of these block diagram devices are highly dependent on the platform on which the embodiments of the present application are to be implemented (that is, these details should be fully within the scope of understanding of those skilled in the art). Where specific details (e.g., circuits) are set forth to describe exemplary embodiments of the present application, it is obvious to those skilled in the art that the embodiments of the present application can be implemented without these specific details or with changes in these specific details. Therefore, these descriptions should be considered illustrative rather than restrictive.

[0124] Although the present application has been described in conjunction with specific embodiments of the present application, many alternatives, modifications and variations of these embodiments will be apparent to those skilled in the art from the foregoing description. For example, other memory architectures (e.g., dynamic RAM (DRAM)) may use the discussed embodiments.

[0125] The embodiments of the present application are intended to cover all such substitutions, modifications and variations that fall within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the embodiments of the present application should be included in the scope of protection of the present application.

Claims

1. A method for determining the accuracy of a pavement defect identification result, comprising: Acquire a pavement disease identification result and a pavement disease benchmark result, wherein the pavement disease identification result includes at least one of a repair identification result or a crack identification result, and the benchmark result includes at least one of a repair benchmark result or a crack benchmark result; Calculating a confidence interval of a repair area according to the repair identification result and the repair benchmark result; Based on the crack identification result and the crack benchmark result, calculating a mean value of crack area errors, and using the mean value of crack area errors as a test statistic; Obtaining a pavement damage condition index, and obtaining a first acceptable range of a mean error of an overall diseased area according to the pavement damage condition index; Determine an acceptance domain limit value and a rejection domain limit value based on the confidence interval and the first acceptable range; Based on the acceptance domain boundary value and the rejection domain boundary value, a rejection domain is obtained; In response to the test statistic falling into the rejection region, it is determined that a pavement defect identification result corresponding to the test statistic is inaccurate.

2. The method of claim 1, wherein: The step of calculating the confidence interval of the repair area according to the repair identification result and the repair benchmark result further comprises: Calculating a mean value of a patch area error and a variance of a patch area error according to the patch identification result and the patch benchmark result; The confidence interval is calculated based on the patch area error mean and the patch area error variance.

3. The method of claim 1, wherein: The first acceptable range of the overall disease area error mean obtained according to the pavement damage index further includes: According to the pavement damage condition index, a second acceptable range of pavement damage rate is obtained; The second acceptable range is converted into the first acceptable range.

4. The method of claim 2, wherein: The mean value of the patching area error is expressed as: in, is the patch recognition result of the i-th image, is the inpainting benchmark result of the i-th image, and n is the number of images.

5. The method of claim 2, wherein: The calculating the confidence interval based on the patching area error mean and the patching area error variance further comprises: According to the repair area error variance, an evaluation error is obtained; The confidence interval is calculated based on the evaluation error and the mean of the repair area error.

6. The method of claim 1, wherein: The acceptance domain limit value is expressed as: , The rejection region boundary value is expressed as: , in, is the first acceptable range, is the minimum mean error of the total diseased area, is the maximum overall disease area error mean, is the confidence interval, is the minimum value of the confidence interval, is the maximum value of the confidence interval.

7. The method of claim 1, wherein: The rejection region is expressed as: in, is the acceptance domain limit value, is the rejection region limit value, is the mean error of the crack area, represents the t distribution with n-1 degrees of freedom, S is the sample variance, n is the sample size, The probability of making a Type I error.

8. A device for determining the accuracy of a road surface disease identification result, comprising: an acquisition module configured to acquire a pavement disease identification result and a pavement disease benchmark result, wherein the pavement disease identification result includes at least one of a repair identification result or a crack identification result, and the benchmark result includes at least one of a repair benchmark result or a crack benchmark result; A first calculation module is configured to calculate a confidence interval of a repair area according to the repair identification result and the repair reference result; A second calculation module is configured to calculate a mean value of crack area errors based on the crack identification result and the crack reference result, and use the mean value of crack area errors as a test statistic; A third calculation module is configured to obtain a pavement damage condition index, and obtain a first acceptable range of a mean error of an overall diseased area according to the pavement damage condition index; a fourth calculation module, configured to determine an acceptance domain boundary value and a rejection domain boundary value based on the confidence interval and the first acceptable range; A fifth calculation module is configured to obtain a rejection domain based on the acceptance domain boundary value and the rejection domain boundary value; The discrimination module is configured to determine that a pavement damage identification result corresponding to the test statistic is inaccurate in response to the test statistic falling into the rejection domain.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the method according to any one of claims 1 to 7 is implemented.

10. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to enable a computer to execute the method according to any one of claims 1 to 7.

Citation Information

Patent Citations

  • Calibration method of automatic recognition accuracy of corrupted data based on pavement images

    CN101923651A

  • Slow disk detection method and device, electronic equipment and readable storage medium

    CN115098314A

  • Automated hypothesis testing

    US20070239361A1