Inspection Method and Device for Culvert

Abstract

The present application relates to the technical field of image processing, and provides a detection method and device for culverts. The method includes: according to the matching result between the rainfall information in the area where the culvert is located and the preset rainfall information, when the rainfall information meets the preset rainfall information, input the inner wall image into the first detection model for image detection, or when the rainfall information does not meet the preset rainfall information, input the inner wall image into a second detection model larger than the first detection model for image detection, so as to obtain the detection result of the culvert. The detection method for culverts provided by the embodiments of the present application can improve the detection efficiency of culverts.

Description

Technical Field

[0001] This application relates to the field of image processing technology, and in particular, to a method and device for detecting culverts. Background Art

[0002] A culvert is a small drainage structure built under a roadbed and is one of the components of a drainage system. To improve the safety of the drainage system, it is usually necessary to detect whether there are defects in the culvert, such as whether there are cracks.

[0003] In the related art, for the detection method of whether there are defects in a culvert, usually, an inner wall image of the culvert is collected and input into a trained image detection model to determine whether there are defects in the culvert.

[0004] To improve the accuracy of culvert defect detection, in the related art, a lightweight image detection model is used to perform a preliminary screening on the inner wall image of the culvert, and then the inner wall image passing the preliminary screening is input into a complex image detection model for further verification to determine whether there are defects in the culvert. However, this detection method may require two detection models to detect the inner wall image, resulting in a large amount of computing resources being occupied, a slow detection speed, and affecting the detection efficiency of the culvert. Summary of the Invention

[0005] This application aims to at least solve one of the technical problems existing in the related art. For this reason, this application proposes a method for detecting a culvert, which can improve the detection efficiency of the culvert.

[0006] This application also proposes a device for detecting a culvert.

[0007] This application also proposes an electronic device.

[0008] This application also proposes a computer-readable storage medium.

[0009] According to an embodiment of the first aspect of this application, the method for detecting a culvert includes:

[0010] Obtain rainfall information of the area where the culvert is located;

[0011] According to the matching result between the rainfall information and the preset rainfall information, input at least one frame of the inner wall image of the culvert obtained within a preset time interval into a first detection model or a second detection model for image detection, and obtain the detection result of the culvert;

[0012] Wherein, the preset time interval is the time interval between the time when the rainfall information is obtained this time and the time when the rainfall information is obtained next time;

[0013] If the matching result is that the rainfall information meets the preset rainfall information, the inner wall image is input into the first detection model for image detection; or if the matching result is that the rainfall information does not meet the preset rainfall information, the inner wall image is input into the second detection model for image detection, and the first detection model is smaller than the second detection model.

[0014] According to the matching result of the rainfall information in the area where the culvert is located and the preset rainfall information, when the rainfall information meets the preset rainfall information, the inner wall image is input into the first detection model for image detection; or when the rainfall information does not meet the preset rainfall information, the inner wall image is input into the second detection model larger than the first detection model for image detection to obtain the detection result of the culvert, so as to improve the accuracy of the detection of the inner wall of the culvert as much as possible while reducing the resource occupancy of the detection of the inner wall of the culvert and improving the detection efficiency of the culvert.

[0015] According to an embodiment of the present application, according to the matching result of the rainfall information and the preset rainfall information, at least one frame of the inner wall image of the culvert obtained within a preset time interval is input into the first detection model or the second detection model for image detection to obtain the detection result of the culvert, including:

[0016] Determine that the rainfall information meets the preset rainfall information, and within the preset time interval, according to the target acquisition frequency greater than the preset acquisition frequency, acquire each of the inner wall images of the culvert;

[0017] Input each of the inner wall images into the first detection model for image detection to obtain the detection result of the culvert.

[0018] According to an embodiment of the present application, inputting each of the inner wall images into the first detection model for image detection to obtain the detection result of the culvert includes:

[0019] Input any target image in each of the inner wall images into the first detection model for image detection to obtain the detection result of the target image;

[0020] Determine that the detection result of the target image is that there is no inner wall defect, and the detection result obtained by inputting the previous frame of the inner wall image of the target image into the first detection model is that there is no inner wall defect, and determine that the detection result of the culvert at the target moment when the target image is acquired is that there is no inner wall defect.

[0021] According to an embodiment of the present application, inputting each of the inner wall images into the first detection model for image detection to obtain the detection result of the culvert includes:

[0022] Input any target image in each of the inner wall images into the first detection model for image detection to obtain the detection result of the target image;

[0023] It is determined that the detection result of the target image is that there is an inner wall defect, and the detection result obtained by inputting the inner wall image of the next frame of the target image into the first detection model is that there is an inner wall defect. It is determined that the detection result of the culvert at the target time when the target image is acquired is that there is an inner wall defect.

[0024] According to an embodiment of the present application, according to the matching result between the rainfall information and the preset rainfall information, at least one inner wall image of the culvert acquired within a preset time interval is input into the first detection model or the second detection model for image detection to obtain the detection result of the culvert, including:

[0025] It is determined that the rainfall information does not meet the preset rainfall information. Within the preset time interval, according to the preset acquisition frequency, each of the inner wall images of the culvert is acquired;

[0026] Each of the inner wall images is input into the second detection model for image detection to obtain the detection result of the culvert.

[0027] According to an embodiment of the present application, it further includes:

[0028] When the rainfall amount in the rainfall information reaches the preset rainfall amount in the preset rainfall information, or the rainfall speed in the rainfall information reaches the preset rainfall speed in the preset rainfall information, it is determined that the rainfall information meets the preset rainfall information.

[0029] According to an embodiment of the present application, it further includes:

[0030] When the rainfall amount in the rainfall information is less than the preset rainfall amount in the preset rainfall information, and the rainfall speed in the rainfall information is less than the preset rainfall speed in the preset rainfall information, it is determined that the rainfall information does not meet the preset rainfall information.

[0031] The culvert detection device according to the second aspect embodiment of the present application includes:

[0032] An information acquisition module, configured to acquire rainfall information of the area where the culvert is located;

[0033] A culvert detection module, configured to input at least one inner wall image of the culvert acquired within a preset time interval into the first detection model or the second detection model for image detection according to the matching result between the rainfall information and the preset rainfall information to obtain the detection result of the culvert;

[0034] Wherein, the preset time interval is the time interval between the time when the rainfall information is acquired this time and the time when the rainfall information is acquired next time;

[0035] If the matching result is that the rainfall information meets the preset rainfall information, input the inner wall image into the first detection model for image detection; or if the matching result is that the rainfall information does not meet the preset rainfall information, input the inner wall image into the second detection model for image detection, where the first detection model is smaller than the second detection model.

[0036] An electronic device according to an embodiment of the third aspect of the present application includes a processor and a memory storing a computer program. When the processor executes the computer program, the detection method of the culvert described in any of the above embodiments is implemented.

[0037] A computer-readable storage medium according to an embodiment of the fourth aspect of the present application stores a computer program thereon. When the computer program is executed by a processor, the detection method of the culvert described in any of the above embodiments is implemented. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the technical solutions in the present application or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0039] Figure 1 It is a flowchart of the detection method of the culvert according to some embodiments of the present application;

[0040] Figure 2 It is a schematic structural diagram of the detection device of the culvert provided by some embodiments of the present application;

[0041] Figure 3 It is a schematic structural diagram of the electronic device provided by some embodiments of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0042] To make the objectives, technical solutions, and advantages of the present application clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are some, but not all, of the embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present application.

[0043] Next, the detection method and device of the culvert provided by the embodiments of the present application will be introduced and described in detail through several specific embodiments.

[0044] In one embodiment, a method for detecting a culvert is provided. This method is applied to a terminal device and is used to detect abnormalities in the culvert. Among them, the terminal device can be a desktop terminal, a mobile terminal or a server. The server can be an independent server or a server cluster composed of multiple servers, and can also be a cloud server that provides basic cloud computing services such as cloud services, cloud message databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN, and large message data and artificial intelligence sampling point devices.

[0045] As Figure 1 shown, a method for detecting a culvert provided in this embodiment includes:

[0046] Step 101, obtain rainfall information of the area where the culvert is located;

[0047] Step 102, according to the matching result between the rainfall information and the preset rainfall information, input at least one frame of inner wall image of the culvert obtained within a preset time interval into the first detection model or the second detection model for image detection, and obtain the detection result of the culvert;

[0048] Among them, the preset time interval is the time interval between the moment when the rainfall information is obtained this time and the moment when the rainfall information is obtained next time;

[0049] If the matching result is that the rainfall information meets the preset rainfall information, input the inner wall image into the first detection model for image detection, or if the matching result is that the rainfall information does not meet the preset rainfall information, input the inner wall image into the second detection model for image detection. The first detection model is smaller than the second detection model.

[0050] In some embodiments, when it is necessary to detect a culvert, rainfall information of the area where the culvert is located can be obtained at a preset time interval. Among them, the preset time interval can be set according to the actual situation. For example, obtain rainfall information of the area where the culvert is located every 10 minutes. The rainfall information can include at least one of rainfall amount or rainfall speed, and can be obtained through methods such as weather forecast or rain gauge detection.

[0051] After obtaining the rainfall information of the area where the culvert is located, the rainfall information can be compared with the preset rainfall information of the same information type. If the rainfall information is rainfall amount, the preset rainfall information is the preset rainfall amount; if the rainfall information is rainfall speed, the preset rainfall information is the preset rainfall speed. Among them, the preset rainfall information can be specifically set according to the actual situation, such as the critical value that causes leakage on the inner wall where the culvert has defects. Exemplarily, if the rainfall information is rainfall speed, the critical rainfall amount that causes leakage on the inner wall where the culvert has defects can be determined as the preset rainfall amount. This critical rainfall amount indicates that if the rainfall amount reaches this critical rainfall amount, it is very likely to cause leakage on the inner wall where the culvert has defects.

[0052] If the rainfall information meets the preset rainfall information, such as the rainfall amount reaches the preset rainfall amount, and at this time if there are defects on the inner wall of the culvert that cause leakage, the defects on the inner wall of the culvert are very likely to cause the culvert to leak. At this time, it is easier to identify the inner wall defects that cause leakage through the inner wall image. Therefore, within the preset time interval between the moment of obtaining the rainfall information this time and the next moment of obtaining the rainfall information, at least the inner wall image of the culvert can be obtained through the imaging device, and the obtained inner wall image is input into the trained first detection model for image detection.

[0053] Among them, the first detection model is a lightweight image detection model, such as a single-stage detection model like YOLO. The training of the first detection model can be carried out by using each inner wall image with inner wall defects as negative samples and each inner wall image without inner wall defects as positive samples, and then inputting each inner wall image sample composed of positive and negative samples into the first detection model for training to obtain the trained first detection model.

[0054] After obtaining the trained first detection model, if the rainfall information of the area where the culvert is located meets the preset rainfall information, it can be determined at this time that if there are inner wall defects on the inner wall of the culvert, these inner wall defects are likely to cause the culvert to leak, and it is easier to identify the inner wall defects that cause leakage through the inner wall image. Therefore, the inner wall images collected within the preset time interval can be input into the trained first detection model for image detection to obtain the detection result output by the first detection model. If the detection result input to the first detection model is that there are inner wall defects in this inner wall image, it can be determined that the detection result of the culvert is that there are inner wall defects; otherwise, it can be determined that the detection result of the culvert is that there are no inner wall defects. Thus, while improving the accuracy of the detection of the inner wall of the culvert as much as possible, the resource occupancy of the detection of the inner wall of the culvert is reduced.

[0055] If the rainfall information does not meet the preset rainfall information, such as the rainfall amount being less than the preset rainfall amount, even if there are inner wall defects on the inner wall of the culvert that cause leakage at this time, it is very likely that the culvert will not leak. Since the first detection model is a lightweight model, it is difficult to identify subtle inner wall defects that will cause leakage on the inner wall by recognizing the inner wall image through the first detection model. More precise image detection is required to identify such inner wall defects. Therefore, within the preset time interval between the current time of obtaining rainfall information and the next time of obtaining rainfall information, at least the inner wall image of the culvert can be obtained through a camera device, and the obtained inner wall image can be input into a trained second detection model for image detection.

[0056] Among them, the second detection model is a non-lightweight image detection model, such as a multi-stage detection model like R-CNN. For the training of the second detection model, similarly, each inner wall image with inner wall defects can be used as a negative sample, and each inner wall image without inner wall defects can be used as a positive sample. Then, each inner wall image sample composed of positive and negative samples can be input into the second detection model for training to obtain a trained second detection model.

[0057] After obtaining the trained second detection model, if the rainfall information in the area where the culvert is located does not meet the preset rainfall information, it can be determined at this time that even if there are inner wall defects on the inner wall of the culvert, it is difficult to identify the inner wall defects that cause leakage from the inner wall image through a lightweight model. Therefore, the inner wall images collected within the preset time interval can be input into the trained second detection model for image detection to obtain the detection result output by the second detection model. If the detection result input to the second detection model is that there are inner wall defects in the inner wall image, it can be determined that the detection result of the culvert is that there are inner wall defects. Otherwise, it can be determined that the detection result of the culvert is that there are no inner wall defects. Thus, for inner wall images that are difficult to identify, they can be directly detected through the second detection model without the need for preliminary screening through the lightweight first detection model, so as to improve the accuracy of the detection of the inner wall of the culvert as much as possible while reducing the resource occupation of the detection of the inner wall of the culvert.

[0058] By matching the rainfall information in the area where the culvert is located with the preset rainfall information, when the rainfall information meets the preset rainfall information, the inner wall image is input into the first detection model for image detection, or when the rainfall information does not meet the preset rainfall information, the inner wall image is input into the second detection model larger than the first detection model for image detection to obtain the detection result of the culvert, so as to improve the accuracy of the detection of the inner wall of the culvert as much as possible while reducing the resource occupation of the detection of the inner wall of the culvert and improving the detection efficiency of the culvert.

[0059] To improve the timeliness of culvert detection, in some embodiments, according to the matching result of the rainfall information and the preset rainfall information, at least one frame of inner wall image of the culvert obtained within a preset time interval is input into the first detection model or the second detection model for image detection to obtain the detection result of the culvert, including:

[0060] Determine that the rainfall information meets the preset rainfall information, and within the preset time interval, obtain each of the inner wall images of the culvert according to a target acquisition frequency greater than the preset acquisition frequency;

[0061] Input each of the inner wall images into the first detection model for image detection to obtain the detection result of the culvert.

[0062] In some embodiments, if the rainfall information in the area where the culvert is located meets the preset rainfall information, such as the rainfall reaches the preset rainfall amount, at this time, if there are defects on the inner wall of the culvert that cause leakage, then the defects on the inner wall of the culvert are likely to cause the culvert to leak. Therefore, the requirement for the timeliness of culvert detection is higher. Based on this, within a preset time interval, a target acquisition frequency greater than the preset acquisition frequency can be used to obtain the inner wall images of the culvert, so as to collect the inner wall images of the culvert at a shorter time interval.

[0063] After the inner wall images of the culvert are collected, the inner wall images of the culvert can be input into the first detection model for image detection to obtain the detection result of the culvert, so that when the rainfall information meets the preset rainfall information, the inner wall images of the culvert can be detected more frequently and quickly, thereby improving the timeliness and detection efficiency of the inner wall detection of the culvert.

[0064] Considering that the first detection model is a lightweight model and the reliability of its detection result may be poor, therefore, to improve the reliability of the detection result of the culvert obtained by the first detection model, in some embodiments, inputting each of the inner wall images into the first detection model for image detection to obtain the detection result of the culvert includes:

[0065] Input any target image in each of the inner wall images into the first detection model for image detection to obtain the detection result of the target image;

[0066] Determine that the detection result of the target image is that there are no inner wall defects, and the detection result obtained by inputting the previous frame of the inner wall image of the target image into the first detection model is that there are no inner wall defects, and determine that the detection result of the culvert at the target moment when the target image is obtained is that there are no inner wall defects.

[0067] In some embodiments, the target image is any one of the inner wall images. Each inner wall image has a corresponding timestamp, which is the target moment when the inner wall image is acquired. For example, if the moment when a certain inner wall image is acquired is T1, then the target moment corresponding to this inner wall image is T1.

[0068] When using the first detection model for culvert detection, the target image can be input into the first detection model to obtain the detection result of the target image. If the detection result of the target image is that there is no inner wall defect, that is, no image area with inner wall defect is detected from the target image, it means that at this time, through the first detection model, it is detected that the culvert has no inner wall defect at the target moment corresponding to the target image. And since each inner wall image is taken in chronological order, if there is no inner wall defect in the target image, then the previous inner wall image of the target image must also have no inner wall defect. At this time, the detection result obtained by inputting the previous inner wall image of the target image into the first detection model can be obtained, and this detection result can be compared with the detection result obtained by inputting the target image into the first detection model to verify the reliability of the detection result that there is no inner wall defect in the target image. If the detection result obtained by inputting the previous inner wall image into the first detection model is that there is an inner wall defect, it can be determined that the detection result that there is no inner wall defect in the target image is unreliable, and at this time, the detection result of the target image is ignored.

[0069] If the detection result obtained by inputting the previous inner wall image into the first detection model is also that there is no inner wall defect, it can be determined that the detection result that there is no inner wall defect in the target image is reliable, and at this time, it can be determined that the detection result of the culvert at the target moment when the target image is acquired is that there is no inner wall defect.

[0070] In this way, the reliability of the detection result that there is no inner wall defect in the target image can be verified by the detection result obtained by inputting the previous inner wall image into the first detection model, so as to improve the reliability of the detection result of the culvert obtained by the first detection model.

[0071] Alternatively, in some embodiments, inputting each of the inner wall images into the first detection model for image detection to obtain the detection result of the culvert includes:

[0072] Inputting any target image among the inner wall images into the first detection model for image detection to obtain the detection result of the target image;

[0073] Determining that the detection result of the target image is that there is an inner wall defect, and the detection result obtained by inputting the next inner wall image of the target image into the first detection model is that there is an inner wall defect, and determining that the detection result of the culvert at the target moment when the target image is acquired is that there is an inner wall defect.

[0074] In some embodiments, if the detection result of the target image indicates the existence of an inner wall defect, that is, an image area with an inner wall defect is detected from the target image, it means that at this time, through the first detection model, it is detected that the culvert has an inner wall defect at the target time corresponding to the target image. Since the inner wall images are captured in chronological order, if the target image has an inner wall defect, the next frame of the inner wall image of the target image will surely also have an inner wall defect. At this time, the detection result obtained by inputting the next frame of the inner wall image of the target image into the first detection model can be acquired, and this detection result can be compared with the detection result obtained by inputting the target image into the first detection model to verify the reliability of the detection result that the target image has an inner wall defect. If the detection result obtained by inputting the next frame of the inner wall image into the first detection model indicates the non-existence of an inner wall defect, it can be determined that the detection result that the target image has an inner wall defect is unreliable, and at this time, the detection result of the target image is ignored.

[0075] If the detection result obtained by inputting the next frame of the inner wall image into the first detection model also indicates the existence of an inner wall defect, it can be determined that the detection result that the target image has an inner wall defect is reliable, and at this time, it can be determined that the detection result of the culvert at the target time when the target image is acquired indicates the existence of an inner wall defect.

[0076] In this way, the reliability of the detection result that the target image has an inner wall defect can be verified through the detection result obtained by inputting the next frame of the inner wall image into the first detection model, so as to improve the reliability of the detection result of the culvert obtained by the first detection model.

[0077] In some embodiments, according to the matching result between the rainfall information and the preset rainfall information, at least one frame of the inner wall image of the culvert acquired within a preset time interval is input into the first detection model or the second detection model for image detection to obtain the detection result of the culvert, including:

[0078] It is determined that the rainfall information does not meet the preset rainfall information. Within the preset time interval, according to the preset acquisition frequency, each of the inner wall images of the culvert is acquired;

[0079] Each of the inner wall images is input into the second detection model for image detection to obtain the detection result of the culvert.

[0080] In some embodiments, if the rainfall information in the area where the culvert is located does not meet the preset rainfall information, such as the rainfall amount is less than the preset rainfall amount, at this time, even if there is an inner wall defect in the inner wall of the culvert that causes leakage, it is highly likely that the culvert will not leak. At the same time, since the culvert does not leak, the change in the image area corresponding to the inner wall defect is also relatively subtle. Based on this, within the preset time interval, the inner wall images of the culvert can be acquired at a frequency greater than the preset acquisition frequency to perform image acquisition on the inner wall of the culvert at a longer time interval.

[0081] After the inner wall image of the culvert is acquired, the inner wall image of the culvert can be input into the second detection model for image detection to obtain the detection result of the culvert, so that when the rainfall information does not meet the preset rainfall information, the ineffective detection of the inner wall image of the culvert can be reduced, and at the same time, the inner wall image can be detected more accurately by the second detection model, so as to improve the accuracy of the detection of the inner wall of the culvert while reducing the resource occupation of the detection of the inner wall of the culvert.

[0082] To more accurately determine whether the rainfall information in the area where the culvert is located will cause the culvert to leak, in some embodiments, the method further includes:

[0083] When the rainfall amount in the rainfall information reaches the preset rainfall amount in the preset rainfall information, or the rainfall speed in the rainfall information reaches the preset rainfall speed in the preset rainfall information, it is determined that the rainfall information meets the preset rainfall information.

[0084] In some embodiments, the rainfall information includes rainfall amount and rainfall speed, and the preset rainfall information includes preset rainfall amount and preset rainfall speed. If the rainfall amount reaches the preset rainfall amount, or the rainfall speed reaches the preset rainfall speed, it means that the defects in the inner wall of the culvert will cause the culvert to leak, and it is easier to find the inner wall defects causing the leak. At this time, it can be determined that the rainfall information meets the preset rainfall information, so as to perform image detection on the inner wall image of the culvert through the first detection model.

[0085] In some embodiments, if the rainfall amount is less than the preset rainfall amount and the rainfall speed is less than the preset rainfall speed, it means that even if there are inner wall defects such as cracks in the inner wall of the culvert, it is highly unlikely that the culvert will leak, and it is difficult to find the inner wall defects causing the leak. At this time, it can be determined that the rainfall information does not meet the preset rainfall information, so as to perform image detection on the inner wall image of the culvert through the second detection model.

[0086] The detection device of the culvert provided by the present application will be described below, and the detection device of the culvert described below can be mutually referred to the detection method of the culvert described above.

[0087] In one embodiment, as Figure 2 shown, a detection device of a culvert is provided, including:

[0088] An information acquisition module 210, configured to acquire rainfall information in the area where the culvert is located;

[0089] A culvert detection module 220, configured to input at least one frame of inner wall image of the culvert acquired within a preset time interval into the first detection model or the second detection model for image detection according to the matching result between the rainfall information and the preset rainfall information, so as to obtain the detection result of the culvert;

[0090] Wherein, the preset time interval is the time interval between the moment of obtaining the rainfall information this time and the moment of obtaining the rainfall information next time;

[0091] If the matching result is that the rainfall information meets the preset rainfall information, the inner wall image is input into the first detection model for image detection; or if the matching result is that the rainfall information does not meet the preset rainfall information, the inner wall image is input into the second detection model for image detection, and the first detection model is smaller than the second detection model.

[0092] According to the matching result between the rainfall information in the area where the culvert is located and the preset rainfall information, when the rainfall information meets the preset rainfall information, the inner wall image is input into the first detection model for image detection; or when the rainfall information does not meet the preset rainfall information, the inner wall image is input into the second detection model larger than the first detection model for image detection, so as to obtain the detection result of the culvert, thereby improving the accuracy of the detection of the inner wall of the culvert as much as possible while reducing the resource occupation of the detection of the inner wall of the culvert and improving the detection efficiency of the culvert.

[0093] In one embodiment, the culvert detection module 220 is specifically configured to:

[0094] Determine that the rainfall information meets the preset rainfall information, and within the preset time interval, obtain each inner wall image of the culvert according to a target acquisition frequency greater than the preset acquisition frequency;

[0095] Input each inner wall image into the first detection model for image detection to obtain the detection result of the culvert.

[0096] In one embodiment, the culvert detection module 220 is specifically configured to:

[0097] Input any target image in each inner wall image into the first detection model for image detection to obtain the detection result of the target image;

[0098] Determine that the detection result of the target image is that there is no inner wall defect, and the detection result obtained by inputting the previous frame of the inner wall image of the target image into the first detection model is that there is no inner wall defect, and determine that the detection result of the culvert at the target moment when the target image is obtained is that there is no inner wall defect.

[0099] In one embodiment, the culvert detection module 220 is specifically configured to:

[0100] Input any target image in each inner wall image into the first detection model for image detection to obtain the detection result of the target image;

[0101] It is determined that the detection result of the target image indicates the existence of an inner wall defect, and the detection result obtained by inputting the next frame of the inner wall image of the target image into the first detection model also indicates the existence of an inner wall defect. It is determined that the detection result of the culvert at the target time when the target image is acquired is the existence of an inner wall defect.

[0102] In one embodiment, the culvert detection module 220 is specifically configured to:

[0103] Determine that the rainfall information does not meet the preset rainfall information, and within the preset time interval, obtain each of the inner wall images of the culvert according to the preset acquisition frequency;

[0104] Input each of the inner wall images into a second detection model for image detection to obtain the detection result of the culvert.

[0105] In one embodiment, the culvert detection module 220 is further configured to:

[0106] When the rainfall amount in the rainfall information reaches the preset rainfall amount in the preset rainfall information, or the rainfall speed in the rainfall information reaches the preset rainfall speed in the preset rainfall information, it is determined that the rainfall information meets the preset rainfall information.

[0107] In one embodiment, the culvert detection module 220 is further configured to:

[0108] When the rainfall amount in the rainfall information is less than the preset rainfall amount in the preset rainfall information, and the rainfall speed in the rainfall information is less than the preset rainfall speed in the preset rainfall information, it is determined that the rainfall information does not meet the preset rainfall information.

[0109] Figure 3 An example of a schematic physical structure diagram of an electronic device is shown as Figure 3 shown. The electronic device may include: a processor 810, a communication interface 820, a memory 830, and a communication bus 840. Among them, the processor 810, the communication interface 820, and the memory 830 complete mutual communication through the communication bus 840. The processor 810 can call a computer program in the memory 830 to execute the detection method of the culvert, for example, including:

[0110] Obtain the rainfall information of the area where the culvert is located;

[0111] According to the matching result between the rainfall information and the preset rainfall information, input at least one frame of the inner wall image of the culvert obtained within the preset time interval into the first detection model or the second detection model for image detection to obtain the detection result of the culvert;

[0112] Wherein, the preset time interval is the time interval between the moment of obtaining the rainfall information this time and the moment of obtaining the rainfall information next time;

[0113] The matching result is that the rainfall information meets the preset rainfall information, and the inner wall image is input into the first detection model for image detection, or the matching result is that the rainfall information does not meet the preset rainfall information, and the inner wall image is input into the second detection model for image detection, and the first detection model is smaller than the second detection model.

[0114] In addition, when the logical instructions in the above-mentioned memory 830 are implemented in the form of software functional units and sold or used as an independent product, they can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods in various embodiments of the present application. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0115] On the other hand, the embodiments of the present application further provide a storage medium. The storage medium includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the detection method of the culvert provided in the above-mentioned various embodiments, for example, including:

[0116] Obtain the rainfall information of the area where the culvert is located;

[0117] According to the matching result between the rainfall information and the preset rainfall information, input at least one frame of the inner wall image of the culvert obtained within the preset time interval into the first detection model or the second detection model for image detection to obtain the detection result of the culvert;

[0118] Wherein, the preset time interval is the time interval between the moment of obtaining the rainfall information this time and the moment of obtaining the rainfall information next time;

[0119] If the matching result is that the rainfall information meets the preset rainfall information, the inner wall image is input into the first detection model for image detection; or if the matching result is that the rainfall information does not meet the preset rainfall information, the inner wall image is input into the second detection model for image detection, and the first detection model is smaller than the second detection model.

[0120] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without creative effort.

[0121] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course also by hardware. Based on this understanding, the above technical solution, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product, which can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., including several instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0122] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of each embodiment of the present application.

Claims

1. A culvert detection method, characterized in that: include: Get rainfall information in the area where the culvert is located; According to the matching result between the rainfall information and the preset rainfall information, at least one frame of the inner wall image of the culvert acquired within the preset time interval is input into the first detection model or the second detection model for image detection to obtain the detection result of the culvert; The preset time interval is the time interval between the moment of obtaining the rainfall information this time and the moment of obtaining the rainfall information next time; The matching result is that the rainfall information satisfies the preset rainfall information, and the inner wall image is input into the first detection model for image detection; or the matching result is that the rainfall information does not satisfy the preset rainfall information, and the inner wall image is input into the second detection model for image detection, the first detection model is smaller than the second detection model, the first detection model is a lightweight image detection model, and the second detection model is a non-lightweight image detection model.

2. The culvert detection method according to claim 1, characterized in that: According to the matching result between the rainfall information and the preset rainfall information, at least one frame of the inner wall image of the culvert acquired within the preset time interval is input into the first detection model or the second detection model for image detection to obtain the detection result of the culvert, including: Determining that the rainfall information satisfies preset rainfall information, and acquiring the inner wall images of the culvert according to a target acquisition frequency greater than a preset acquisition frequency within the preset time interval; Each of the inner wall images is input into a first detection model for image detection to obtain a detection result of the culvert.

3. The culvert detection method according to claim 2, characterized in that: Inputting each of the inner wall images into a first detection model for image detection to obtain a detection result of the culvert, including: Inputting any target image among the inner wall images into the first detection model for image detection to obtain a detection result of the target image; Determine that the detection result of the target image is that there is no inner wall defect, and the detection result obtained by inputting the inner wall image of the previous frame of the target image into the first detection model is that there is no inner wall defect, and determine that the detection result of the culvert at the target moment of obtaining the target image is that there is no inner wall defect.

4. The culvert detection method according to claim 2, characterized in that: Inputting each of the inner wall images into a first detection model for image detection to obtain a detection result of the culvert, including: Inputting any target image among the inner wall images into the first detection model for image detection to obtain a detection result of the target image; Determine that the detection result of the target image is that there is an inner wall defect, and the detection result obtained by inputting the inner wall image of the next frame of the target image into the first detection model is that there is an inner wall defect, and determine that the detection result of the culvert at the target moment of obtaining the target image is that there is an inner wall defect.

5. The culvert detection method according to claim 1, characterized in that: According to the matching result between the rainfall information and the preset rainfall information, at least one frame of the inner wall image of the culvert acquired within the preset time interval is input into the first detection model or the second detection model for image detection to obtain the detection result of the culvert, including: Determining that the rainfall information does not meet the preset rainfall information, acquiring the inner wall images of the culvert according to the preset acquisition frequency within the preset time interval; Each of the inner wall images is input into a second detection model for image detection to obtain a detection result of the culvert.

6. The culvert detection method according to any one of claims 1 to 5, characterized in that: Also includes: When the rainfall amount in the rainfall information reaches the preset rainfall amount in the preset rainfall information, or the rainfall speed in the rainfall information reaches the preset rainfall speed in the preset rainfall information, it is determined that the rainfall information meets the preset rainfall information.

7. The culvert detection method according to any one of claims 1 to 5, characterized in that: Also includes: If the rainfall amount in the rainfall information is less than the preset rainfall amount in the preset rainfall information, and the rainfall speed in the rainfall information is less than the preset rainfall speed in the preset rainfall information, it is determined that the rainfall information does not meet the preset rainfall information.

8. A culvert detection device, characterized in that: include: An information acquisition module is used to obtain rainfall information in the area where the culvert is located; A culvert detection module, for inputting at least one frame of inner wall image of the culvert acquired within a preset time interval into a first detection model or a second detection model for image detection according to a matching result between the rainfall information and preset rainfall information, so as to obtain a detection result of the culvert; The preset time interval is the time interval between the moment of obtaining the rainfall information this time and the moment of obtaining the rainfall information next time; The matching result is that the rainfall information satisfies the preset rainfall information, and the inner wall image is input into the first detection model for image detection; or the matching result is that the rainfall information does not satisfy the preset rainfall information, and the inner wall image is input into the second detection model for image detection, the first detection model is smaller than the second detection model, the first detection model is a lightweight image detection model, and the second detection model is a non-lightweight image detection model.

9. An electronic device comprising a processor and a memory storing a computer program, characterized in that: When the processor executes the computer program, the culvert detection method according to any one of claims 1 to 7 is implemented.

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

Images