Road construction quality detection device and method

Abstract

The present invention relates to the field of image processing technology, and in particular to a road construction quality detection device and method, comprising: calculating a preset abnormality degree for the jth area of ​​the pth thermal image collected by a thermal imager to obtain a target abnormality probability of the jth area of ​​the pth thermal image; when the target abnormality probability is not less than a preset first threshold, using the target abnormality probability to correct the first gradient threshold of n historical thermal images to obtain a second gradient threshold of the n historical thermal images; determining the jth area of ​​each historical thermal image corresponding to the jth area of ​​the pth thermal image in the n historical thermal images according to the second gradient threshold of the n historical thermal images; determining the road construction quality detection result of the jth area according to the target abnormality probability of the jth area of ​​each historical thermal image and the target abnormality probability of the jth area of ​​the pth thermal image. The present invention effectively improves the accuracy of road construction quality detection.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and in particular to a road construction quality detection device and method. Background Art

[0002] Road construction quality inspection is a key link in ensuring the safety, reliability and economy of road projects. It prevents early road damage, increases road service life, and ensures the driving safety of users by evaluating the quality of construction materials, processes and final products, while also reducing the cost of long-term maintenance and repair. In the existing road construction quality inspection, in order to improve the inspection efficiency and accuracy, a variety of road pavement construction quality inspection methods and systems based on visual technology have been developed to inspect the road pavement construction quality.

[0003] However, most of these methods are based on acquiring road images with ordinary cameras. Ordinary cameras cannot directly capture temperature differences caused by internal defects or unevenness of materials. In addition, ordinary cameras are greatly affected by lighting conditions and may not work effectively at night or in low-light environments. There is often a lot of dust in the air at construction sites, which will affect the image acquisition quality of ordinary cameras, thereby resulting in low accuracy in road construction quality inspection results. Summary of the invention

[0004] In order to solve the technical problem of low accuracy of road construction quality detection results, the purpose of the present invention is to provide a road construction quality detection device and method, and the technical solution adopted is as follows:

[0005] Obtain the pth thermal image of the target construction section collected by the thermal imager;

[0006] Performing a preset abnormality calculation on the j-th area of ​​the p-th thermal image to obtain a target abnormality probability of the j-th area of ​​the p-th thermal image;

[0007] When the target abnormality probability of the j-th area of ​​the p-th thermal image is not less than a preset first threshold, the first gradient threshold of the n historical thermal images before the p-th thermal image is corrected using the target abnormality probability of the j-th area of ​​the p-th thermal image to obtain a second gradient threshold of the n historical thermal images;

[0008] Dividing the n historical thermal images into regions according to the second gradient threshold of the n historical thermal images, and determining the j-th region of each historical thermal image corresponding to the j-th region of the p-th thermal image in the division results;

[0009] Performing the preset abnormality degree calculation on the j-th area of ​​each historical thermal image to obtain the target abnormality probability of the j-th area of ​​each historical thermal image;

[0010] The road construction quality inspection result of the jth area is determined according to the target abnormality probability of the jth area of ​​each historical thermal image and the target abnormality probability of the jth area of ​​the pth thermal image; wherein p, j, and n are all positive integers, and p is greater than n.

[0011] Preferably, the preset abnormality degree calculation includes:

[0012] According to the heat values ​​of the j-th area of ​​the thermal image and its adjacent areas, the target abnormality probability of the j-th area of ​​the thermal image is determined.

[0013] Preferably, determining the target abnormality probability of the jth area of ​​the thermal image according to the heat values ​​of the jth area of ​​the thermal image and its adjacent areas includes:

[0014] Determining a predicted abnormal probability of the jth region of the thermal image according to the heat values ​​of the jth region and its adjacent regions;

[0015] The predicted abnormal probability of the jth area in the thermal image is corrected by using the positional relationship between n historical thermal images before the thermal image and the jth area in the thermal image, so as to obtain the target abnormal probability of the jth area in the thermal image.

[0016] Preferably, determining the predicted abnormal probability of the jth area of ​​the thermal image according to the heat values ​​of the jth area of ​​the thermal image and its adjacent areas includes:

[0017] Determine a rate of change of the thermal value of the jth region of the thermal image according to the thermal value of each pixel in the jth region of the thermal image and the thermal values ​​of the pixels adjacent to the pixel;

[0018] The predicted abnormal probability of the jth area of ​​the thermal image is determined based on the difference between the thermal value average of the jth area of ​​the thermal image and the thermal value average of its adjacent areas, the thermal value variance of the adjacent areas, and the thermal value change rate of the jth area of ​​the thermal image.

[0019] Preferably, the method for determining the positional relationship between n historical thermal images before the thermal image and the j-th area of ​​the thermal image comprises:

[0020] Determine the area in each historical thermal image that is closest to the center point of the jth area of ​​the thermal image as the target area of ​​each historical thermal image;

[0021] Connect the center points of the target area of ​​each historical thermal image in chronological order, and connect the center point of the jth area of ​​the thermal image with the center point of the previous historical thermal image to form n vectors;

[0022] The angles between adjacent vectors in the n vectors are summed to obtain a first angle;

[0023] The vector length of each vector in the n vectors and the similarity between adjacent thermal images are summed to obtain a similarity parameter;

[0024] According to the product of the first angle and the similarity parameter, the positional relationship between n historical thermal images before the thermal image and the j-th area of ​​the thermal image is obtained.

[0025] Preferably, the method for determining the first gradient threshold of n historical thermal images before the pth thermal image comprises:

[0026] The target area of ​​each historical thermal image is compared with the target area of ​​the previous historical thermal image to obtain n change areas;

[0027] The mean value of the gradient of each pixel point in the n change areas is calculated respectively, and the first gradient threshold of the n historical thermal images is obtained respectively.

[0028] Preferably, the first gradient threshold of n historical thermal images before the pth thermal image is corrected using the target abnormality probability of the jth area of ​​the pth thermal image to obtain the second gradient threshold of the n historical thermal images, including:

[0029] The target abnormality probability of each area of ​​the p-th thermal image is multiplied by the first gradient threshold of the n historical thermal images respectively to obtain the predicted gradient threshold of the n historical thermal images;

[0030] The predicted gradient threshold of the n historical thermal images is corrected according to a preset correction coefficient to obtain a second gradient threshold of the n historical thermal images; wherein the preset correction coefficient is determined according to the number of regions in the pth thermal image.

[0031] Preferably, determining the road construction quality inspection result of the jth area according to the target abnormality probability of the jth area of ​​each historical thermal image and the target abnormality probability of the jth area of ​​the pth thermal image includes:

[0032] The maximum probability value among the target abnormality probability of the j-th area of ​​each historical thermal image and the target abnormality probability of the j-th area of ​​the p-th thermal image is determined as the road construction quality detection result of the j-th area.

[0033] Preferably, the method further comprises:

[0034] When the road construction quality detection result of the j-th area is not less than a preset second threshold, road abnormality information for the j-th area is sent to the staff.

[0035] The present invention also provides a road construction quality detection device, comprising:

[0036] An acquisition module is used to acquire the pth thermal image of the target construction section collected by the thermal imager;

[0037] A first processing module, used for calculating a preset abnormality degree of the jth region of the pth thermal image to obtain a target abnormality probability of the jth region of the pth thermal image;

[0038] A second processing module is used to correct the first gradient threshold of n historical thermal images before the p-th thermal image using the target abnormality probability of the j-th area of ​​the p-th thermal image when the target abnormality probability of the j-th area of ​​the p-th thermal image is not less than a preset first threshold, so as to obtain a second gradient threshold of the n historical thermal images;

[0039] a third processing module, configured to divide the n historical thermal images into regions according to the second gradient threshold of the n historical thermal images, and determine the j-th region of each historical thermal image corresponding to the j-th region of the p-th thermal image in the division result;

[0040] A calculation module, used for performing the preset abnormality degree calculation on the j-th area of ​​each historical thermal image to obtain the target abnormality probability of the j-th area of ​​each historical thermal image;

[0041] The detection module is used to determine the road construction quality detection result of the jth area according to the target abnormality probability of the jth area of ​​each historical thermal image and the target abnormality probability of the jth area of ​​the pth thermal image; wherein p, j, and n are all positive integers, and p is greater than n.

[0042] The present invention has the following beneficial effects:

[0043] The p-th thermal image of the target construction section collected by the thermal imager is obtained; a preset abnormality degree is calculated for the j-th area of ​​the p-th thermal image to obtain the target abnormality probability of the j-th area of ​​the p-th thermal image; when the target abnormality probability of the j-th area of ​​the p-th thermal image is not less than the preset first threshold, the first gradient threshold of the n historical thermal images before the p-th thermal image is corrected using the target abnormality probability of the j-th area of ​​the p-th thermal image to obtain the second gradient threshold of the n historical thermal images; the n historical thermal images are respectively divided into regions according to the second gradient threshold of the n historical thermal images, and the j-th area of ​​each historical thermal image corresponding to the j-th area of ​​the p-th thermal image is determined in the division result; the preset abnormality degree is calculated for the j-th area of ​​each historical thermal image to obtain the target abnormality probability of the j-th area of ​​each historical thermal image; the road construction quality inspection result of the j-th area is determined according to the target abnormality probability of the j-th area of ​​each historical thermal image and the target abnormality probability of the j-th area of ​​the p-th thermal image. In this way, the thermal imager is used to obtain images of the target construction section that needs to be inspected for quality, and the road construction quality inspection result of the jth area is judged by combining the abnormality degree of the pth thermal image and its historical thermal image, thereby realizing non-destructive inspection of various areas of the road, thereby improving the accuracy of road construction quality inspection. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0045] Figure 1 A method flow chart of a road construction quality detection method provided by an embodiment of the present invention;

[0046] Figure 2 A flowchart of a method for calculating a preset abnormality degree in a road construction quality detection method provided by an embodiment of the present invention;

[0047] Figure 3 A specific method flow chart of step S130 in a road construction quality detection method provided by an embodiment of the present invention;

[0048] Figure 4 A schematic diagram of a road construction quality detection device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0049] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of a road construction quality detection device and method proposed by the present invention, its specific implementation, structure, features and effects, in conjunction with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0050] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0051] The specific scheme of a road construction quality detection device and method provided by the present invention is described in detail below with reference to the accompanying drawings.

[0052] See also Figure 1 , which shows a method flow chart of a road construction quality detection method provided by an embodiment of the present invention. In an exemplary embodiment, a road construction quality detection method is provided, comprising:

[0053] S110, obtaining the pth thermal image of the target construction section collected by the thermal imager;

[0054] S120, performing a preset abnormality degree calculation on the j-th area of ​​the p-th thermal image to obtain a target abnormality probability of the j-th area of ​​the p-th thermal image;

[0055] S130, when the target abnormality probability of the j-th area of ​​the p-th thermal image is not less than a preset first threshold, the first gradient threshold of n historical thermal images before the p-th thermal image is corrected by using the target abnormality probability of the j-th area of ​​the p-th thermal image to obtain a second gradient threshold of the n historical thermal images;

[0056] S140, dividing the n historical thermal images into regions according to the second gradient threshold of the n historical thermal images, and determining the j-th region of each historical thermal image corresponding to the j-th region of the p-th thermal image in the division results;

[0057] S150, performing the preset abnormality degree calculation on the j-th area of ​​each historical thermal image to obtain the target abnormality probability of the j-th area of ​​each historical thermal image;

[0058] S160. Determine the road construction quality inspection result of the jth area according to the target abnormality probability of the jth area of ​​each historical thermal image and the target abnormality probability of the jth area of ​​the pth thermal image; wherein p, j, and n are all positive integers, and p is greater than n.

[0059] In step S110, illustratively, the target construction section may be a designated or arbitrarily selected section, which is not limited here. The target construction section may be one section or multiple sections. The division of the construction section is artificially divided according to actual conditions. The pth thermal image may be an image collected at a designated time, or an image fed back in real time by a thermal imager.

[0060] Optionally, one or more thermal imagers may be installed in each construction section. The specific installation method of the thermal imager requires first understanding the layout, length, width and characteristics of the construction section, determining the key areas that need to be detected, confirming the detection range, resolution, accuracy and working conditions of the selected thermal imager, and ensuring that it meets the monitoring needs of the construction section. Relevant personnel shall formulate a detailed monitoring plan, which includes the layout of monitoring points, monitoring frequency and monitoring time period, and install thermal imagers at monitoring points determined in advance, adjust the angle and direction, ensure that the field of view of the thermal imager can cover the predetermined monitoring area, and set an overlapping area of ​​a certain width between adjacent monitoring points to avoid monitoring blind spots. During the process of the thermal imager monitoring the section to which it belongs, the thermal image of the monitored section is transmitted back to the data processing center every 10 minutes for storage, and the road surface during the construction process is continuously monitored.

[0061] In step S120, illustratively, the jth region of the pth thermal image may be any region in the pth thermal image, or may be one or more designated regions, which are not limited here. Specifically, the pth thermal image is divided into regions in advance, and the preset abnormality degree of each region in the pth thermal image is calculated to obtain the target abnormality probability of each region. Among them, the specific division rule of the thermal image is based on the heat value in the collected thermal image, and the canny edge detection algorithm is adopted, in which the upper and lower thresholds are set to 30 and 10 respectively. When the initial edge detection threshold is small, the number of edges obtained is large, and the acquired thermal image is divided into regions in combination with the contour extraction algorithm.

[0062] Preferably, the preset abnormality degree calculation includes: determining the target abnormality probability of the jth area of ​​the thermal image according to the heat values ​​of the jth area of ​​the thermal image and its adjacent areas.

[0063] For example, for thermal images of road construction quality inspection, irregular operations during the construction process, such as uneven paving, inconsistent compaction, etc., may lead to uneven heat distribution. Defects such as moisture, cracks or voids that may exist inside the road surface will also affect the transfer and distribution of heat. Therefore, the area indicating the abnormality often has a large difference in temperature from other surrounding areas, and its heat distribution pattern is also quite different from that of adjacent areas. In abnormal areas, the rate of change of heat values ​​may appear abnormally rapid or abnormally slow compared to normal areas. It should be noted that the preset abnormality degree calculation is applicable to any thermal image. Therefore, the heat value of the jth area of ​​the pth thermal image is compared with the heat value of its adjacent area to obtain the rate of change of the heat value, so as to judge the target abnormality probability of the jth area of ​​the pth thermal image based on the rate of change of the heat value.

[0064] Furthermore, if Figure 2 As shown, according to the heat values ​​of the j-th area of ​​the thermal image and its adjacent areas, the target abnormality probability of the j-th area of ​​the thermal image is determined, including: S210, according to the heat values ​​of the j-th area of ​​the thermal image and its adjacent areas, the predicted abnormality probability of the j-th area of ​​the thermal image is determined; S220, using the positional relationship between n historical thermal images before the thermal image and the j-th area of ​​the thermal image, the predicted abnormality probability of the j-th area in the thermal image is corrected to obtain the target abnormality probability of the j-th area of ​​the thermal image.

[0065] Exemplarily, n historical thermal images and the thermal image are images taken by the same thermal imager, and n is set according to actual needs. In this embodiment, it is set to 5. It can be understood that there are 5 images before the pth thermal image.

[0066] Optionally, step S210 includes: determining a rate of change of a heat value of the jth region of the thermal image according to heat values ​​of each pixel point in the jth region of the thermal image and heat values ​​of pixels adjacent to the pixel points;

[0067] The predicted abnormal probability of the jth area of ​​the thermal image is determined based on the difference between the thermal value average of the jth area of ​​the thermal image and the thermal value average of its adjacent areas, the thermal value variance of the adjacent areas, and the thermal value change rate of the jth area of ​​the thermal image.

[0068] Specifically, taking the jth region of the pth thermal image as an example, set In the area, take Pixel points are counted and the sum of the heat value change rates of the point compared to other points in the 3×3 neighborhood is calculated. The sum of the change rates is divided by the number of calculations of the heat value change rate, and the average change rate is recorded as , No. There are a total of pixels, and the The adjacent area The variance of the heat values ​​of all points in the region is The average heat value of this area is The average difference in heat between the regions is , a total of Region and The regions are adjacent, then The probability that a region represents an abnormal region It can be expressed as:

[0069]

[0070] in, Represents the rate of change of the heat value of the jth area of ​​the thermal image. It should be noted that in the above formula, when the variance of the heat value in a certain area is smaller, the area can better represent the normal area, and the difference in heat between the area to be sought and the area can better represent the abnormal probability of the area. However, in the process of area division, some shadow areas blocked by other objects or people may be identified as separate areas, and these areas may also have the characteristics of different temperatures from normal areas due to being blocked. If the shadow area is identified as an abnormal area, it will affect the overall abnormality assessment. Since most shadow areas do not exist continuously, the abnormal areas appear more stably in multiple images, and their shapes are also more stable. When abnormal areas appear in multiple images, their positions and shapes should be similar, and the persistent shadows generally move in a fixed direction with the change of the sun's position.

[0071] Optionally, a method for determining a positional relationship between n historical thermal images before the thermal image and the j-th area of ​​the thermal image comprises:

[0072] Determine the area in each historical thermal image that is closest to the center point of the jth area of ​​the thermal image as the target area of ​​each historical thermal image;

[0073] Connect the center points of the target area of ​​each historical thermal image in chronological order, and connect the center point of the jth area of ​​the thermal image with the center point of the previous historical thermal image to form n vectors;

[0074] The angles between adjacent vectors in the n vectors are summed to obtain a first angle;

[0075] The vector length of each vector in the n vectors and the similarity between adjacent thermal images are summed to obtain a similarity parameter;

[0076] According to the product of the first angle and the similarity parameter, the positional relationship between n historical thermal images before the thermal image and the j-th area of ​​the thermal image is obtained.

[0077] Specifically, taking the jth region of the pth thermal image and n=5 as an example, take the first five images of the current latest image and set them in the p-1th thermal image (i.e., the historical thermal image) and the The area of ​​the region closest to the center point is , the distance between the center points of the two regions is recorded as , by In the thermal image The center point of the corresponding area of ​​the pth image is the starting point. The center point of the corresponding area of ​​the region is the end point, and the vector formed is . and The angle between Understandably, and The construction method is the same as that of the p-th thermal image. The thermal images before the p-th thermal image are processed in sequence according to the above method. Then, the SIFT algorithm is used to extract the p-1-th thermal image and the p-2-th thermal image. The key points of the target area corresponding to the pth area, and the pth thermal image extracted by SIFT algorithm The key points corresponding to the regions are obtained, and a feature descriptor is generated for each key point. The best matching feature point pair is found by calculating the cosine similarity between the descriptors. When the similarity is greater than 0.8, the feature points are considered to match each other. The number of matching feature point pairs in the two regions is obtained as follows: , based on this The probability that a region represents an abnormal region Correction, the probability after correction It can be expressed as:

[0078]

[0079] In the above formula, Represents the similarity parameter, that is, in the five images before the current image, The corresponding similar area that can be found in the area is The similarity between the two regions. The higher the similarity, the more it means that the two regions represent the same region in actual situation. It represents the first angle, which indicates the degree to which the center point of the same area in multiple images moves in one direction. The more uniform the moving direction is, the more likely it is that the area is a shadow area (i.e., a non-abnormal area).

[0080] In step S130, illustratively, the preset first threshold is set according to actual conditions and is not limited here. In this embodiment, the preset first threshold is 0.64. If the target abnormal probability of the jth region of the pth thermal image is not less than 0.64, the jth region is regarded as an abnormal region. And according to the pth region in the latest image, The abnormality of the abnormal area is determined by the performance of each area and the changes in the area in the previous five images. When the temperature difference of a certain area is greater than that of the normal area and the area is larger, the abnormality of the area is higher. At the same time, when its area gradually increases over time, the abnormality of the area is higher.

[0081] Since the above region segmentation results are not accurate enough, only the position information of each region is preliminarily obtained. The optimal gradient threshold corresponding to each image is obtained below, and then edge detection is performed to extract the defect area corresponding to each frame image.

[0082] Optionally, the method for determining the first gradient threshold of n historical thermal images before the pth thermal image includes: comparing the target area of ​​each historical thermal image with the target area of ​​the previous historical thermal image to obtain n change areas; calculating the mean of the gradient of each pixel point in the n change areas respectively, and obtaining the first gradient threshold of the n historical thermal images respectively.

[0083] Specifically, take the first five thermal images before the pth thermal image (i.e., historical thermal images). In any historical thermal image, the pth The area corresponding to the region of the image that changes compared with the area corresponding to the previous image is recorded as ,in , in the corresponding In the region, the mean gradient of each point in the previous frame is calculated. The gradient mean refers to the mean of the gradients of all points in an area, and its calculation method can be to use the frame difference to obtain the change area, and the gradient mean of all points in the change area is used as the first gradient threshold of the historical thermal image.

[0084] Furthermore, if Figure 3 As shown, step S130 includes: S1310, the target abnormal probability of each area of ​​the p-th thermal image is multiplied by the first gradient threshold of n historical thermal images respectively to obtain the predicted gradient threshold of the n historical thermal images;

[0085] S1320. Correct the predicted gradient threshold of the n historical thermal images according to a preset correction coefficient to obtain a second gradient threshold of the n historical thermal images; wherein the preset correction coefficient is determined according to the number of regions in the pth thermal image.

[0086] Specifically, since the change area is an area with obvious edge changes, the gradient mean in this area is used as the final edge detection threshold, which can better segment different frame images. Therefore, the second gradient threshold T of the historical thermal image is recorded as:

[0087]

[0088] in, Indicates the total number of regions obtained by segmenting the historical thermal image. Represents the first gradient threshold of the historical thermal image. The higher the probability that a certain area represents an abnormal area, the more likely it is that the area represents an abnormal area. At this time, the gradient of the change area of ​​the corresponding area in multiple images can better represent the edge gradient of the abnormal area. When calculating the gradient threshold of the entire image, the higher the average gradient of the change area corresponding to the area is given. It should be noted that, taking n as 5 as an example, for each historical thermal image, the corresponding second gradient threshold is calculated, and 5 second gradient thresholds can be obtained.

[0089] Furthermore, when the target abnormality probability of the j-th area of ​​the p-th thermal image is less than a preset first threshold, it is determined that the j-th area of ​​the p-th thermal image is not an abnormal area, and there is no need to determine the area corresponding to the j-th area of ​​the p-th thermal image in the n historical thermal images.

[0090] In step S140, for example, each historical thermal image is divided into regions, and the second gradient threshold of each historical thermal image is used as the upper threshold for canny edge detection, and half of the second gradient threshold is used as the lower threshold for canny edge detection. In this way, the corresponding edge detection results in multiple historical images can be obtained respectively, and then the corresponding multiple abnormal regions in the historical thermal images (i.e., the jth region of each historical thermal image) can be obtained respectively using the traditional contour extraction algorithm. In this way, multiple regions corresponding to the jth region of the pth thermal image are obtained, so that the overall abnormality of the monitored road section can be judged according to the abnormality of the jth region of multiple historical thermal images.

[0091] In step S150, illustratively, a preset abnormality degree calculation is used to calculate the corresponding multiple abnormal areas in the historical thermal images (i.e., the jth area of ​​each historical thermal image). Specifically, the predicted abnormal probability of the jth area of ​​each historical thermal image is determined based on the heat value of the jth area of ​​each historical thermal image and its adjacent areas; the predicted abnormal probability of the jth area of ​​each historical thermal image is corrected by using the positional relationship between the n historical thermal images before each historical thermal image and the jth area of ​​the thermal image, and the target abnormal probability of the jth area of ​​each historical thermal image is obtained. In this embodiment, the use and The corresponding formula is calculated for the j-th area of ​​each historical thermal image to obtain the target abnormality probability of the j-th area of ​​each historical thermal image.

[0092] In step S160, the road construction quality detection result indicates the abnormal probability of the jth area of ​​the target road section. Specifically, different weights may be set for different thermal images, and the abnormal probability of each thermal image target is weightedly calculated according to the weights to obtain the abnormal probability of the jth area.

[0093] Preferably, step S160 includes: determining the maximum probability value of the target abnormality probability of the jth area of ​​each historical thermal image and the target abnormality probability of the jth area of ​​the pth thermal image as the road construction quality inspection result of the jth area. The specific formula of the road construction quality inspection result of the jth area can be expressed as: .

[0094] In the technical solution of the present application, the pth thermal image of the target construction section collected by the thermal imager is obtained; the preset abnormality degree is calculated for the jth area of ​​the pth thermal image to obtain the target abnormality probability of the jth area of ​​the pth thermal image; when the target abnormality probability of the jth area of ​​the pth thermal image is not less than the preset first threshold, the first gradient threshold of n historical thermal images before the pth thermal image is corrected using the target abnormality probability of the jth area of ​​the pth thermal image to obtain the second gradient threshold of the n historical thermal images; the n historical thermal images are respectively divided into regions according to the second gradient threshold of the n historical thermal images, and the jth area of ​​each historical thermal image corresponding to the jth area of ​​the pth thermal image is determined in the division result; the preset abnormality degree is calculated for the jth area of ​​each historical thermal image to obtain the target abnormality probability of the jth area of ​​each historical thermal image; the road construction quality inspection result of the jth area is determined according to the target abnormality probability of the jth area of ​​each historical thermal image and the target abnormality probability of the jth area of ​​the pth thermal image. In this way, the thermal imager is used to obtain images of the target construction section that needs to be inspected for quality, and the road construction quality inspection result of the jth area is judged by combining the abnormality degree of the pth thermal image and its historical thermal image, thereby realizing non-destructive inspection of various areas of the road, thereby improving the accuracy of road construction quality inspection.

[0095] In one embodiment, the method further comprises:

[0096] When the road construction quality detection result of the jth area is not less than the preset second threshold, the road abnormality information for the jth area is sent to the staff. The preset second threshold is set according to the actual situation and is not limited here. In this embodiment, the preset second threshold can be 0.75.

[0097] Specifically, according to the road construction quality inspection result of the jth area, when the road construction quality inspection result of the jth area When the road anomaly information for the jth area is sent to the staff through SMS or email, the relevant personnel are reminded to use more accurate means to detect the area, clarify the specific problems corresponding to the area, and explore improvement measures in combination with the existing process flow, so as to effectively solve the problem of road anomalies.

[0098] See also Figure 4 , which shows a schematic diagram of a road construction quality detection device provided by an embodiment of the present invention. In an exemplary embodiment, a road construction quality detection device is provided, comprising:

[0099] An acquisition module 410 is used to acquire a p-th thermal image of the target construction section captured by a thermal imager;

[0100] A first processing module 420 is used to calculate a preset abnormality degree for the j-th area of ​​the p-th thermal image to obtain a target abnormality probability for the j-th area of ​​the p-th thermal image;

[0101] The second processing module 430 is used to correct the first gradient threshold of n historical thermal images before the p-th thermal image using the target abnormality probability of the j-th area of ​​the p-th thermal image when the target abnormality probability of the j-th area of ​​the p-th thermal image is not less than the preset first threshold, so as to obtain the second gradient threshold of the n historical thermal images; p is greater than n;

[0102] A third processing module 440 is used to divide the n historical thermal images into regions according to the second gradient threshold of the n historical thermal images, and determine the j-th region of each historical thermal image corresponding to the j-th region of the p-th thermal image in the division result;

[0103] A calculation module 450 is used to perform the preset abnormality degree calculation on the j-th area of ​​each historical thermal image to obtain a target abnormality probability of the j-th area of ​​each historical thermal image;

[0104] The detection module 460 is used to determine the road construction quality detection result of the jth area according to the target abnormality probability of the jth area of ​​each historical thermal image and the target abnormality probability of the jth area of ​​the pth thermal image.

[0105] The road construction quality detection device provided in this embodiment belongs to the same application concept as the road construction quality detection method provided in the above embodiment of this application, and can execute the road construction quality detection method provided in any of the above embodiments of this application, and has the corresponding functional modules and beneficial effects of executing a road construction quality detection method. For technical details not described in detail in this embodiment, please refer to the specific processing content of the road construction quality detection method provided in the above embodiment of this application, and will not be repeated here.

[0106] In other embodiments, a device is also provided, including a memory and a processor. The memory is used to store executable program code, and the processor is used to call and run the executable program code from the memory, so that the device executes the above-mentioned road construction quality detection device and method. The device can be specifically a chip, a component or a module, and the chip may include a connected processor and memory; wherein the memory is used to store instructions, and when the processor calls and executes the instructions, the chip can execute the road construction quality detection device and method provided in the above-mentioned embodiment.

[0107] In other embodiments, a computer program product is also provided. When the computer program product is run on a computer, the computer is enabled to execute the above-mentioned related steps to implement a road construction quality detection method provided in the above-mentioned embodiment.

[0108] In other embodiments, a computer-readable storage medium is also provided, in which a computer program code is stored. When the computer program code is executed on a computer, the computer executes the above-mentioned related method steps to implement a road construction quality detection method provided in the above-mentioned embodiment.

[0109] Among them, the provided systems, devices, computer program products, and computer-readable storage media are all used to execute the corresponding methods provided above. Therefore, the beneficial effects that can be achieved can refer to the beneficial effects in the corresponding methods provided above and will not be repeated here.

[0110] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. 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.

[0111] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A road construction quality detection method, characterized in that: The method comprises: Obtain the pth thermal image of the target construction section collected by the thermal imager; Performing a preset abnormality calculation on the j-th area of ​​the p-th thermal image to obtain a target abnormality probability of the j-th area of ​​the p-th thermal image; When the target abnormality probability of the j-th area of ​​the p-th thermal image is not less than a preset first threshold, the first gradient threshold of the n historical thermal images before the p-th thermal image is corrected using the target abnormality probability of the j-th area of ​​the p-th thermal image to obtain a second gradient threshold of the n historical thermal images; Dividing the n historical thermal images into regions according to the second gradient threshold of the n historical thermal images, and determining the j-th region of each historical thermal image corresponding to the j-th region of the p-th thermal image in the division results; Performing the preset abnormality degree calculation on the j-th area of ​​each historical thermal image to obtain the target abnormality probability of the j-th area of ​​each historical thermal image; The road construction quality inspection result of the jth area is determined according to the target abnormality probability of the jth area of ​​each historical thermal image and the target abnormality probability of the jth area of ​​the pth thermal image; wherein p, j, and n are all positive integers, and p is greater than n.

2. The road construction quality detection method according to claim 1, characterized in that: The preset abnormality degree calculation methods include: According to the heat values ​​of the j-th area of ​​the thermal image and its adjacent areas, the target abnormality probability of the j-th area of ​​the thermal image is determined.

3. The road construction quality detection method according to claim 2 is characterized in that: Determining the target abnormality probability of the jth area of ​​the thermal image according to the heat values ​​of the jth area of ​​the thermal image and its adjacent areas includes: Determining a predicted abnormal probability of the jth region of the thermal image according to the heat values ​​of the jth region and its adjacent regions; The predicted abnormal probability of the jth area in the thermal image is corrected by using the positional relationship between n historical thermal images before the thermal image and the jth area in the thermal image, so as to obtain the target abnormal probability of the jth area in the thermal image.

4. The road construction quality detection method according to claim 3 is characterized in that: Determining the predicted abnormal probability of the jth area of ​​the thermal image according to the heat values ​​of the jth area of ​​the thermal image and its adjacent areas, including: Determining a rate of change of the thermal value of the jth region of the thermal image according to the thermal values ​​of each pixel in the jth region of the thermal image and the thermal values ​​of its adjacent pixels; The predicted abnormal probability of the jth area of ​​the thermal image is determined based on the difference between the thermal value average of the jth area of ​​the thermal image and the thermal value average of its adjacent areas, the thermal value variance of the adjacent areas, and the thermal value change rate of the jth area of ​​the thermal image.

5. The road construction quality detection method according to claim 3 is characterized in that: The method for determining the positional relationship between n historical thermal images before the thermal image and the j-th area of ​​the thermal image comprises: Determine the area in each historical thermal image that is closest to the center point of the jth area of ​​the thermal image as the target area of ​​each historical thermal image; Connect the center points of the target area of ​​each historical thermal image in chronological order, and connect the center point of the jth area of ​​the thermal image with the center point of the previous historical thermal image to form n vectors; The angles between adjacent vectors in the n vectors are summed to obtain a first angle; The vector length of each vector in the n vectors and the similarity between adjacent thermal images are summed to obtain a similarity parameter; According to the product of the first angle and the similarity parameter, the positional relationship between n historical thermal images before the thermal image and the j-th area of ​​the thermal image is obtained.

6. The road construction quality detection method according to claim 5, characterized in that: The method for determining the first gradient threshold of n historical thermal images before the pth thermal image comprises: The target area of ​​each historical thermal image is compared with the target area of ​​the previous historical thermal image to obtain n change areas; The mean value of the gradient of each pixel point in the n change areas is calculated respectively, and the first gradient threshold of the n historical thermal images is obtained respectively.

7. The road construction quality detection method according to claim 6, characterized in that: Using the target abnormality probability of the jth area of ​​the pth thermal image to correct the first gradient threshold of the n historical thermal images before the pth thermal image to obtain the second gradient threshold of the n historical thermal images, including: The target abnormality probability of each area of ​​the p-th thermal image is multiplied by the first gradient threshold of the n historical thermal images to obtain the predicted gradient threshold of the n historical thermal images; The predicted gradient threshold of the n historical thermal images is corrected according to a preset correction coefficient to obtain a second gradient threshold of the n historical thermal images; wherein the preset correction coefficient is determined according to the number of regions in the pth thermal image.

8. The road construction quality detection method according to claim 1, characterized in that: Determining the road construction quality inspection result of the jth area according to the target abnormality probability of the jth area of ​​each historical thermal image and the target abnormality probability of the jth area of ​​the pth thermal image, including: The maximum probability value among the target abnormality probability of the j-th area of ​​each historical thermal image and the target abnormality probability of the j-th area of ​​the p-th thermal image is determined as the road construction quality detection result of the j-th area.

9. The road construction quality detection method according to claim 8, characterized in that: The method further comprises: When the road construction quality detection result of the j-th area is not less than a preset second threshold, road abnormality information for the j-th area is sent to the staff.

10. A road construction quality detection device, characterized in that: include: An acquisition module is used to acquire the pth thermal image of the target construction section collected by the thermal imager; A first processing module, used for calculating a preset abnormality degree of the jth region of the pth thermal image to obtain a target abnormality probability of the jth region of the pth thermal image; A second processing module is used to correct the first gradient threshold of n historical thermal images before the p-th thermal image using the target abnormality probability of the j-th area of ​​the p-th thermal image when the target abnormality probability of the j-th area of ​​the p-th thermal image is not less than a preset first threshold, so as to obtain a second gradient threshold of the n historical thermal images; a third processing module, configured to divide the n historical thermal images into regions according to the second gradient threshold of the n historical thermal images, and determine the j-th region of each historical thermal image corresponding to the j-th region of the p-th thermal image in the division result; A calculation module, used for performing the preset abnormality degree calculation on the j-th area of ​​each historical thermal image to obtain the target abnormality probability of the j-th area of ​​each historical thermal image; The detection module is used to determine the road construction quality detection result of the jth area according to the target abnormality probability of the jth area of ​​each historical thermal image and the target abnormality probability of the jth area of ​​the pth thermal image; wherein p, j, and n are all positive integers, and p is greater than n.

Images