CN117197774B - Road feature acquisition method and device, computer equipment and storage medium - Google Patents
Road feature acquisition method and device, computer equipment and storage medium Download PDFInfo
- Publication number
- CN117197774B CN117197774B CN202310900214.3A CN202310900214A CN117197774B CN 117197774 B CN117197774 B CN 117197774B CN 202310900214 A CN202310900214 A CN 202310900214A CN 117197774 B CN117197774 B CN 117197774B
- Authority
- CN
- China
- Prior art keywords
- image
- information
- target
- dimensional
- damage
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Landscapes
- Analysing Materials By The Use Of Radiation (AREA)
Abstract
The invention relates to a road feature acquisition method, a road feature acquisition device, a computer device and a storage medium. The method comprises the following steps: acquiring image information of a target road, spatial structure information of the target road and radiowave detection information of the target road; identifying each defect characteristic image of the image information and the image range of each defect characteristic image through a self-attention mechanism, and identifying each abnormal three-dimensional fluctuation range of the target road based on the radiation wave detection information and the spatial structure information of the target road; in the space structure information, determining each target damage area of the target road, and determining damage information corresponding to each target damage area based on image information of the target road; and extracting characteristic information of each target damage area based on the damage information of each target damage area, and taking the characteristic information of all the target damage areas as target characteristic information of a target road. By adopting the method, the accuracy of the obtained road characteristics can be improved.
Description
Technical Field
The present invention relates to the field of road reconstruction and expansion technologies, and in particular, to a method and apparatus for obtaining road characteristics, a computer device, and a storage medium.
Background
Along with the continuous and rapid increase of the road traffic, and the restriction of various factors such as the original construction condition, construction concept, road condition, maintenance requirement, service requirement and the like, traffic jam and road damage phenomena are generated, the normal service level and traffic safety of the road are affected, and the problems of expansion and capacity expansion of the expressway are gradually highlighted. The highway extension needs to acquire and analyze the space information of the current continental to make an extension scheme, so that how to quickly and accurately acquire the space information of the existing road is the research focus of the current highway extension.
The traditional road space information acquisition mode is obtained through manual field measurement or based on laser radar point cloud manual pickup. And this mode needs to throw in a large amount of manpower resources, and not only measuring cost is thrown in high, work efficiency is low, and when the manual work picks up laser point cloud data acquisition road characteristic moreover, the personnel subjective randomness is big to lead to the accurate lower of the road characteristic of acquireing.
Disclosure of Invention
The technical problem to be solved by the present invention is to provide a road feature acquisition method, apparatus, computer device, computer readable storage medium and computer program product, which address the above-mentioned shortcomings of the prior art.
The technical scheme for solving the technical problems is as follows:
in a first aspect, the present application provides a road feature acquisition method. The method comprises the following steps:
acquiring image information of a target road, spatial structure information of the target road and radiation wave detection information of the target road;
identifying each defect characteristic image of the image information and an image range of each defect characteristic image through a self-attention mechanism, and identifying each abnormal three-dimensional fluctuation range of the target road based on the radiowave detection information and the spatial structure information of the target road;
Determining each target damage area of the target road in the spatial structure information based on each abnormal three-dimensional fluctuation range and each image range of the defect characteristic image, and determining damage information corresponding to each target damage area based on the image information of the target road;
And extracting characteristic information of each target damage area based on the damage information of each target damage area, and taking the characteristic information of all the target damage areas as target characteristic information of the target road.
Optionally, the identifying each defect feature image of the image information and the image range of each defect feature image by a self-attention mechanism includes:
inputting the image information into a self-attention network to obtain each initial defect image in the image information, and identifying the image characteristics of each initial defect image through an image characteristic identification network;
Taking initial defect images corresponding to image features meeting defect feature information as defect images, and based on the position information of each defect image in the image information of the target road, performing splicing processing on the defect images connected with the position information to obtain each defect feature image;
and identifying the image range of each defect characteristic image based on the image information.
Optionally, the identifying, based on the radiowave detection information and the spatial structure information of the target road, the different three-dimensional fluctuation ranges of the target road includes:
Projecting the radiation wave detection information into the space structure information to obtain a three-dimensional structure fluctuation image of the target road;
Identifying each abnormal structure fluctuation image in the three-dimensional structure fluctuation image, and identifying the image edge of each abnormal structure fluctuation image through an image edge algorithm;
and taking the range included by the image edge of each abnormal structure fluctuation image as an abnormal three-dimensional fluctuation range.
Optionally, the determining, in the spatial structure information, each target damaged area of the target road based on each abnormal three-dimensional fluctuation range and each image range of the defect feature image includes:
performing dimension reduction processing on the three-dimensional structure fluctuation image to obtain a two-dimensional structure image of the three-dimensional structure fluctuation image, fusing the two-dimensional structure image with the image information, and judging whether a sub-two-dimensional structure image corresponding to an abnormal three-dimensional fluctuation range intersected with the defect characteristic image exists or not;
Under the condition that a sub-two-dimensional structural image corresponding to an abnormal three-dimensional fluctuation range intersected with a defect characteristic image exists, carrying out image range adjustment processing on the two-dimensional image range corresponding to the sub-two-dimensional structural image intersected with the defect characteristic image based on the image range of each defect characteristic image to obtain each two-dimensional target damage area, and projecting each two-dimensional target damage area into the three-dimensional structural fluctuation image to obtain each target damage area of the target road;
under the condition that sub two-dimensional structure images corresponding to abnormal three-dimensional fluctuation ranges intersecting with the defect feature images do not exist, projecting the image ranges of the defect feature images into the space structure information, identifying defect feature three-dimensional structure information corresponding to each defect feature image, and taking the three-dimensional image ranges corresponding to the defect feature three-dimensional structure information as a first target damage area of the target road;
And regarding the three-dimensional image range of each abnormal three-dimensional fluctuation range in the space structure information as a second target damage area, and regarding all first target damage areas and all second target damage areas as each target damage area of the target road.
Optionally, the determining, based on the image information of the target road, damage information corresponding to each target damage area includes:
Performing dimension reduction treatment on each target damage region to obtain two-dimensional target damage regions corresponding to each target damage region, and inquiring damage ranges corresponding to each two-dimensional target damage region in the image information;
and identifying the damage type of each damage range through an image identification strategy, and taking the damage range containing the damage type as damage information corresponding to each target damage area.
Optionally, the extracting feature information of each target damage area based on the damage information of each target damage area includes:
Extracting image features of each target damage region through a feature recognition strategy, and adding feature labels to the image features based on the damage range of each target damage region and the damage type of each damage range to obtain feature information of each target damage region.
In a second aspect, the application further provides a road feature acquisition device. The device comprises:
The acquisition module is used for acquiring image information of a target road, spatial structure information of the target road and radiation wave detection information of the target road;
The identification module is used for identifying each defect characteristic image of the image information and the image range of each defect characteristic image through a self-attention mechanism, and identifying each abnormal three-dimensional fluctuation range of the target road based on the radiation wave detection information and the spatial structure information of the target road;
The determining module is used for determining each target damage area of the target road in the space structure information based on each abnormal three-dimensional fluctuation range and the image range of each defect characteristic image, and determining damage information corresponding to each target damage area based on the image information of the target road;
the extraction module is used for extracting the characteristic information of each target damage area based on the damage information of each target damage area, and taking the characteristic information of all the target damage areas as the target characteristic information of the target road.
Optionally, the identification module is specifically configured to:
inputting the image information into a self-attention network to obtain each initial defect image in the image information, and identifying the image characteristics of each initial defect image through an image characteristic identification network;
Taking initial defect images corresponding to image features meeting defect feature information as defect images, and based on the position information of each defect image in the image information of the target road, performing splicing processing on the defect images connected with the position information to obtain each defect feature image;
and identifying the image range of each defect characteristic image based on the image information.
Optionally, the identification module is specifically configured to:
Projecting the radiation wave detection information into the space structure information to obtain a three-dimensional structure fluctuation image of the target road;
Identifying each abnormal structure fluctuation image in the three-dimensional structure fluctuation image, and identifying the image edge of each abnormal structure fluctuation image through an image edge algorithm;
and taking the range included by the image edge of each abnormal structure fluctuation image as an abnormal three-dimensional fluctuation range.
Optionally, the determining module is specifically configured to:
performing dimension reduction processing on the three-dimensional structure fluctuation image to obtain a two-dimensional structure image of the three-dimensional structure fluctuation image, fusing the two-dimensional structure image with the image information, and judging whether a sub-two-dimensional structure image corresponding to an abnormal three-dimensional fluctuation range intersected with the defect characteristic image exists or not;
Under the condition that a sub-two-dimensional structural image corresponding to an abnormal three-dimensional fluctuation range intersected with a defect characteristic image exists, carrying out image range adjustment processing on the two-dimensional image range corresponding to the sub-two-dimensional structural image intersected with the defect characteristic image based on the image range of each defect characteristic image to obtain each two-dimensional target damage area, and projecting each two-dimensional target damage area into the three-dimensional structural fluctuation image to obtain each target damage area of the target road;
under the condition that sub two-dimensional structure images corresponding to abnormal three-dimensional fluctuation ranges intersecting with the defect feature images do not exist, projecting the image ranges of the defect feature images into the space structure information, identifying defect feature three-dimensional structure information corresponding to each defect feature image, and taking the three-dimensional image ranges corresponding to the defect feature three-dimensional structure information as a first target damage area of the target road;
And regarding the three-dimensional image range of each abnormal three-dimensional fluctuation range in the space structure information as a second target damage area, and regarding all first target damage areas and all second target damage areas as each target damage area of the target road.
Optionally, the determining module is specifically configured to:
Performing dimension reduction treatment on each target damage region to obtain two-dimensional target damage regions corresponding to each target damage region, and inquiring damage ranges corresponding to each two-dimensional target damage region in the image information;
and identifying the damage type of each damage range through an image identification strategy, and taking the damage range containing the damage type as damage information corresponding to each target damage area.
Optionally, the extracting module is specifically configured to:
Extracting image features of each target damage region through a feature recognition strategy, and adding feature labels to the image features based on the damage range of each target damage region and the damage type of each damage range to obtain feature information of each target damage region.
In a third aspect, the present application provides a computer device. The computer device comprises a memory storing a computer program and a processor implementing the steps of the method of any of the first aspects when the processor executes the computer program.
In a fourth aspect, the present application provides a computer-readable storage medium. On which a computer program is stored which, when being executed by a processor, implements the steps of the method of any of the first aspects.
In a fifth aspect, the present application provides a computer program product. The computer program product comprising a computer program which, when executed by a processor, implements the steps of the method of any of the first aspects.
The road characteristic acquisition method, the road characteristic acquisition device, the computer equipment and the storage medium are used for acquiring the image information of the target road, the space structure information of the target road and the radiation wave detection information of the target road; identifying each defect characteristic image of the image information and an image range of each defect characteristic image through a self-attention mechanism, and identifying each abnormal three-dimensional fluctuation range of the target road based on the radiowave detection information and the spatial structure information of the target road; determining each target damage area of the target road in the spatial structure information based on each abnormal three-dimensional fluctuation range and each image range of the defect characteristic image, and determining damage information corresponding to each target damage area based on the image information of the target road; and extracting characteristic information of each target damage area based on the damage information of each target damage area, and taking the characteristic information of all the target damage areas as target characteristic information of the target road. The image range of the defect characteristic image of the target road and the abnormal three-dimensional fluctuation range of the target road are determined through the image information of the target road and the radiation wave detection information of the target road, the image ranges and the three-dimensional fluctuation ranges are subjected to range optimization processing through the space structure information of the target road to obtain target damage areas, and finally the damage information of each target damage area is identified, so that the characteristic information of each target damage area is extracted, the surface damage characteristics of each road structure can be accurately identified, the deep damage characteristics of each road structure can be identified, and the accuracy of the obtained road characteristics is improved.
Drawings
FIG. 1 is a flow chart of a road feature acquisition method according to an embodiment;
FIG. 2 is a block diagram of a road feature acquisition device in one embodiment;
FIG. 3 is an internal block diagram of a computer device in one embodiment.
Detailed Description
The present application will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present application more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the application.
The road characteristic acquisition method provided by the embodiment of the application can be applied to a terminal, a server and a system comprising the terminal and the server, and is realized through interaction of the terminal and the server. The terminal may include, but is not limited to, various personal computers, notebook computers, tablet computers, and the like. The terminal determines the image range of the defect characteristic image of the target road and the abnormal three-dimensional fluctuation range of the target road through the image information of the target road and the radiation wave detection information of the target road, performs range optimization processing on the image ranges and the three-dimensional fluctuation ranges through the spatial structure information of the target road to obtain each target damage area, and finally identifies the damage information of each target damage area, so that the characteristic information of each target damage area is extracted, the surface damage characteristics of each road structure can be accurately identified, the deep damage characteristics of each road structure can be identified, and the accuracy of the obtained road characteristics is improved.
In one embodiment, as shown in fig. 1, a road feature obtaining method is provided, and the method is applied to a terminal for illustration, and includes the following steps:
Step S101, acquiring image information of a target road, spatial structure information of the target road, and radiowave detection information of the target road.
In this embodiment, the terminal scans the target road through a radiation wave detection program preset in the terminal to obtain three-dimensional detection information of the target road, and uses the three-dimensional detection information as radiation wave detection information of the target road. And then the terminal acquires image information of the target road through each camera device preset around the target road, wherein the image information is a two-dimensional image. The radiation wave detection procedure may be, but is not limited to, radar, vibration wave detector, etc. And finally, the terminal screens the space three-dimensional structure image of the target road from the road database to obtain the space structure information of the target road.
Step S102, identifying each defect characteristic image of the image information and the image range of each defect characteristic image through a self-attention mechanism, and identifying each abnormal three-dimensional fluctuation range of the target road based on the radiation wave detection information and the spatial structure information of the target road.
In this embodiment, the terminal inputs the image information of the target road into the self-attention network after training is completed, and identifies the defect feature image of the target road. The defect feature image is an image corresponding to a surface defect of the target road, and the surface defect may be, but is not limited to, a crack, a collapse, a breakage, a bulge, a fault, and the like. The terminal identifies the image edge of each defect feature image to determine the image range of each defect feature image, wherein the identification mode of the image edge is identified through an image edge identification algorithm, and the edge identification algorithm comprises, but is not limited to, an algorithm corresponding to Roberts operator, an algorithm corresponding to Prewi tt operator, an algorithm corresponding to Sobe l operator, an algorithm corresponding to Canny operator, an algorithm corresponding to LAP L AC I AN operator and the like. The specific process of identifying the defect-feature image will be described in detail later. The terminal projects the radiation wave detection information of the target road to the space structure information of the target road to obtain image information of the radiation wave detection information, and then identifies a range corresponding to an abnormal image in the image information to obtain various abnormal three-dimensional fluctuation ranges. Wherein the image information is a three-dimensional image. The specific identification process will be described in detail later. The abnormal fluctuation three-dimensional range is used for representing the surface defect and the deep defect of the target road.
Step S103, determining each target damage area of the target road in the space structure information based on each abnormal three-dimensional fluctuation range and the image range of each defect characteristic image, and determining damage information corresponding to each target damage area based on the image information of the target road.
In this embodiment, the terminal determines each target damaged area of the target road in the spatial structure information based on each abnormal three-dimensional fluctuation range and the image range of each defect feature image. Then, the terminal determines damage information corresponding to each target damage region based on the image information of the target road. The damage information is a damage image of the target damage area and damage types corresponding to the damage images. The specific determination will be described in detail later.
Step S104, extracting the characteristic information of each target damage area based on the damage information of each target damage area, and taking the characteristic information of all the target damage areas as the target characteristic information of the target road.
In this embodiment, the terminal extracts the feature information of each target damage area through an image feature extraction algorithm, and obtains the target feature information of each target damage area. The image feature extraction algorithm is a deep reinforcement learning-based graph neural network.
Based on the scheme, the image range of the defect characteristic image of the target road and the abnormal three-dimensional fluctuation range of the target road are determined through the image information of the target road and the radiation wave detection information of the target road, the image ranges and the three-dimensional fluctuation ranges are subjected to range optimization processing through the spatial structure information of the target road to obtain each target damage area, and finally the damage information of each target damage area is identified, so that the characteristic information of each target damage area is extracted, the surface damage characteristics of each road structure can be accurately identified, the deep damage characteristics of each road structure can be identified, and the accuracy of the obtained road characteristics is improved.
Optionally, identifying each defect feature image of the image information and the image range of each defect feature image by a self-attention mechanism includes: inputting the image information into a self-attention network to obtain each initial defect image in the image information, and identifying the image characteristics of each initial defect image through an image characteristic identification network; taking an initial defect image corresponding to the image characteristics meeting the defect characteristic information as a defect image, and based on the position information of the image information of each defect image on a target road, performing splicing processing on the defect images connected with the position information to obtain each defect characteristic image; based on the image information, an image range of each defect feature image is identified.
In this embodiment, the terminal inputs the image information of the target road into the self-attention network to obtain each initial defect image in the image information. Wherein the range size of each initial defect image is the same. For example, images of 4cm x 4cm each. Then, the terminal recognizes image features of each initial defective image through an image feature recognition network. Wherein the image feature recognition network may be, but is not limited to, an image feature extraction network (VGG 16) based on a plurality of convolution layers and a full-pooling layer.
The terminal judges whether each image feature meets the defect feature information or not, and takes an initial defect image corresponding to the image feature meeting the defect feature information as a defect image. The defect characteristic information is characteristic information of various types of surface road defects of the extracted staff experience, the Internet road defect information data and the road defect information in the history database, wherein the characteristic information is preset in the terminal. Then, the terminal performs image stitching processing on the defect images connected with the position information based on the position information of the image information of each defect image on the target road, and obtains each defect characteristic image. Then, the terminal recognizes the image edge of each defect image in the image information by an image edge recognition algorithm, and takes the range included in the image edge of each defect feature image as the image range of each defect feature image.
Based on the scheme, the defect characteristic image is identified through the image information, so that the accuracy of the identified surface defects is improved.
Optionally, identifying the abnormal three-dimensional fluctuation range of the target road based on the spatial structure information of the target road includes: projecting the radial wave detection information into the space structure information to obtain a three-dimensional structure fluctuation image of the target road; identifying each abnormal structure fluctuation image in the three-dimensional structure fluctuation image, and identifying the image edge of each abnormal structure fluctuation image through an image edge algorithm; the range included in the image edge of each abnormal structure fluctuation image is taken as an abnormal three-dimensional fluctuation range.
In this embodiment, the terminal projects the radiation wave detection information to the spatial structure information to obtain the three-dimensional structure fluctuation image of the target road. The three-dimensional structure fluctuation image is used for representing abnormal fluctuation information of the target road in the space structure, and each abnormal fluctuation information is used for reflecting one depth defect. Then, the terminal recognizes each abnormal structure fluctuation image in the three-dimensional structure fluctuation image by an image feature recognition algorithm. The image feature recognition algorithm is a three-dimensional feature description algorithm (3D-WHGO) of the three-dimensional image feature recognition algorithm, the terminal recognizes the image edge of each abnormal structure fluctuation image through an image edge algorithm, and finally, the terminal takes the range included by the image edge of each abnormal structure fluctuation image as an abnormal three-dimensional fluctuation range.
Based on the scheme, the terminal projects the radiation wave detection information to the space structure information, so that the terminal is different from each abnormal three-dimensional fluctuation range, and the accuracy of the identified abnormal three-dimensional fluctuation range is improved.
Optionally, determining each target damaged area of the target road in the spatial structure information based on each abnormal three-dimensional fluctuation range and the image range of each defect feature image includes: performing dimension reduction processing on the three-dimensional structure fluctuation image to obtain a two-dimensional structure image of the three-dimensional structure fluctuation image, fusing the two-dimensional structure image with image information, and judging whether a sub-two-dimensional structure image corresponding to an abnormal three-dimensional fluctuation range intersected with the defect characteristic image exists or not; under the condition that sub-two-dimensional structural images corresponding to abnormal three-dimensional fluctuation ranges intersected with the defect characteristic images exist, based on the image ranges of the defect characteristic images, performing image range adjustment processing on the two-dimensional image ranges corresponding to the sub-two-dimensional structural images intersected with the defect characteristic images to obtain two-dimensional target damage areas, and projecting the two-dimensional target damage areas into the three-dimensional structural fluctuation images to obtain target damage areas of a target road; under the condition that sub two-dimensional structure images corresponding to abnormal three-dimensional fluctuation ranges intersecting the defect feature images do not exist, projecting the image ranges of the defect feature images into the space structure information, identifying defect feature three-dimensional structure information corresponding to each defect feature image, and taking the three-dimensional image range corresponding to the defect feature three-dimensional structure information as a first target damage area of a target road; and regarding the three-dimensional image range of each abnormal three-dimensional fluctuation range in the spatial structure information as a second target damage area, and regarding all the first target damage areas and all the second target damage areas as each target damage area of the target road.
In this embodiment, the terminal maps the three-dimensional structure fluctuation image from three dimensions to a two-dimensional plane, and obtains a two-dimensional structure image of the three-dimensional structure fluctuation image. Then, the terminal performs image fusion processing on the two-dimensional structure image and the image information to obtain a two-dimensional structure image of the three-dimensional structure fluctuation image in the image information of the target road. And the terminal judges whether a sub-two-dimensional structure image corresponding to the abnormal three-dimensional fluctuation range intersected with the defect characteristic image exists or not. When a sub-two-dimensional structure image corresponding to an abnormal three-dimensional fluctuation range intersecting the defect feature image exists, the terminal performs image range adjustment processing on the two-dimensional image range corresponding to each sub-two-dimensional structure image intersecting the defect feature image based on the image range of each defect feature image, and takes the intersecting part of the defect feature image and the sub-two-dimensional structure image as a two-dimensional target damage area. Then, the terminal projects each two-dimensional target damage area into the original three-dimensional structure fluctuation image to obtain each target damage area of the target road. And under the condition that the sub-two-dimensional structural image corresponding to the abnormal three-dimensional fluctuation range intersected with the defect characteristic image does not exist, the terminal projects the image range of each defect characteristic image into the spatial structural information, and the defect characteristic three-dimensional structural information corresponding to each defect characteristic image is identified. And then the terminal takes the three-dimensional image range corresponding to the defect characteristic three-dimensional structure information as a first target damage area of the target road. The terminal uses the three-dimensional image range of each abnormal three-dimensional fluctuation range in the space structure information as a second target damage area, and uses all the first target damage areas and all the second target damage areas as each target damage area of the target road. And when the sub-two-dimensional structure image corresponding to the abnormal three-dimensional fluctuation range which is intersected with the defect characteristic image exists and the sub-two-dimensional structure image corresponding to the abnormal three-dimensional fluctuation range which is not intersected with the defect characteristic image exists, respectively processing the images corresponding to different judging results according to the image processing modes of different judging structure results to obtain each target damage area.
Based on the scheme, the terminal obtains each target damage area by processing the images corresponding to different judging results according to the image processing modes of different judging structure results, and the determination accuracy of each target damage area is improved while the global property of the obtained target damage area is ensured.
Optionally, determining the damage information corresponding to each target damage area based on the image information of the target road includes: performing dimension reduction treatment on each target damage region to obtain two-dimensional target damage regions corresponding to each target damage region, and inquiring damage ranges corresponding to each two-dimensional target damage region in image information; and identifying the damage type of each damage range through an image identification strategy, and taking the damage range containing the damage type as damage information corresponding to each target damage area.
In this embodiment, the terminal reduces the dimension of each target damage region from the three-dimensional image to the two-dimensional image, and obtains a two-dimensional target damage region corresponding to each target damage region. And then, the terminal inquires the damage range corresponding to each two-dimensional target damage area in the image information of the target road. The terminal identifies the damage type of each damage range through an image identification strategy, and takes the damage range containing the damage type as damage information corresponding to each target damage area. The damage types comprise surface damage and depth damage, the type of the damage range is determined to be the depth damage when no surface defect is identified in each identified damage range, and the type of the damage range is determined to be the surface defect when the surface defect is identified in each identified damage range.
Based on the scheme, the efficiency of extracting the characteristic information later is improved by carrying out rules on each target damage area.
Optionally, extracting feature information of each target damage area based on damage information of each target damage area includes: and extracting image features of each target damage region through a feature recognition strategy, and adding feature labels to each image feature based on the damage range of each target damage region and the damage type of each damage range to obtain feature information of each target damage region.
In this embodiment, the terminal directly extracts the two-dimensional image features of the target damaged area by an image feature extraction strategy for the target damaged area with the type of surface damage. Aiming at target damage areas with the types of depth damage, the terminal extracts three-dimensional image features of each target damage area through a three-dimensional feature recognition strategy. Then, the terminal obtains feature information of each target damage region by taking the damage range of the target damage region and the damage type of the damage range as feature tags of image features of the target damage region for each target damage region.
Based on the scheme, the efficiency of extracting the characteristic information is improved by classifying and extracting the target damage areas of different types.
It should be understood that, although the steps in the flowcharts related to the above embodiments are sequentially shown as indicated by arrows, these steps are not necessarily sequentially performed in the order indicated by the arrows. The steps are not strictly limited to the order of execution unless explicitly recited herein, and the steps may be executed in other orders. Moreover, at least some of the steps in the flowcharts described in the above embodiments may include a plurality of steps or a plurality of stages, which are not necessarily performed at the same time, but may be performed at different times, and the order of the steps or stages is not necessarily performed sequentially, but may be performed alternately or alternately with at least some of the other steps or stages.
Based on the same inventive concept, the embodiment of the application also provides a road feature acquisition device for realizing the road feature acquisition method. The implementation of the solution provided by the device is similar to the implementation described in the above method, so the specific limitation in the embodiments of the road feature obtaining device or devices provided below may refer to the limitation of the road feature obtaining method hereinabove, and will not be described herein.
In one embodiment, as shown in fig. 2, there is provided a road feature acquisition apparatus including: an acquisition module 210, an identification module 220, a determination module 230, and an extraction module 240, wherein:
An acquiring module 210, configured to acquire image information of a target road, spatial structure information of the target road, and radiowave detection information of the target road;
An identifying module 220, configured to identify each defect feature image of the image information and an image range of each defect feature image by a self-attention mechanism, and identify each abnormal three-dimensional fluctuation range of the target road based on the radiowave detection information and the spatial structure information of the target road;
a determining module 230, configured to determine, in the spatial structure information, each target damaged area of the target road based on each abnormal three-dimensional fluctuation range and an image range of each defect feature image, and determine damage information corresponding to each target damaged area based on image information of the target road;
the extracting module 240 is configured to extract feature information of each target damaged area based on the damage information of each target damaged area, and use the feature information of all target damaged areas as target feature information of the target road.
Optionally, the identifying module 220 is specifically configured to:
inputting the image information into a self-attention network to obtain each initial defect image in the image information, and identifying the image characteristics of each initial defect image through an image characteristic identification network;
Taking initial defect images corresponding to image features meeting defect feature information as defect images, and based on the position information of each defect image in the image information of the target road, performing splicing processing on the defect images connected with the position information to obtain each defect feature image;
and identifying the image range of each defect characteristic image based on the image information.
Optionally, the identifying module 220 is specifically configured to:
Projecting the radiation wave detection information into the space structure information to obtain a three-dimensional structure fluctuation image of the target road;
Identifying each abnormal structure fluctuation image in the three-dimensional structure fluctuation image, and identifying the image edge of each abnormal structure fluctuation image through an image edge algorithm;
and taking the range included by the image edge of each abnormal structure fluctuation image as an abnormal three-dimensional fluctuation range.
Optionally, the determining module 230 is specifically configured to:
performing dimension reduction processing on the three-dimensional structure fluctuation image to obtain a two-dimensional structure image of the three-dimensional structure fluctuation image, fusing the two-dimensional structure image with the image information, and judging whether a sub-two-dimensional structure image corresponding to an abnormal three-dimensional fluctuation range intersected with the defect characteristic image exists or not;
Under the condition that a sub-two-dimensional structural image corresponding to an abnormal three-dimensional fluctuation range intersected with a defect characteristic image exists, carrying out image range adjustment processing on the two-dimensional image range corresponding to the sub-two-dimensional structural image intersected with the defect characteristic image based on the image range of each defect characteristic image to obtain each two-dimensional target damage area, and projecting each two-dimensional target damage area into the three-dimensional structural fluctuation image to obtain each target damage area of the target road;
under the condition that sub two-dimensional structure images corresponding to abnormal three-dimensional fluctuation ranges intersecting with the defect feature images do not exist, projecting the image ranges of the defect feature images into the space structure information, identifying defect feature three-dimensional structure information corresponding to each defect feature image, and taking the three-dimensional image ranges corresponding to the defect feature three-dimensional structure information as a first target damage area of the target road;
And regarding the three-dimensional image range of each abnormal three-dimensional fluctuation range in the space structure information as a second target damage area, and regarding all first target damage areas and all second target damage areas as each target damage area of the target road.
Optionally, the determining module 230 is specifically configured to:
Performing dimension reduction treatment on each target damage region to obtain two-dimensional target damage regions corresponding to each target damage region, and inquiring damage ranges corresponding to each two-dimensional target damage region in the image information;
and identifying the damage type of each damage range through an image identification strategy, and taking the damage range containing the damage type as damage information corresponding to each target damage area.
Optionally, the extracting module 240 is specifically configured to:
Extracting image features of each target damage region through a feature recognition strategy, and adding feature labels to the image features based on the damage range of each target damage region and the damage type of each damage range to obtain feature information of each target damage region.
The respective modules in the road feature acquisition apparatus described above may be implemented in whole or in part by software, hardware, or a combination thereof. The above modules may be embedded in hardware or may be independent of a processor in the computer device, or may be stored in software in a memory in the computer device, so that the processor may call and execute operations corresponding to the above modules.
In one embodiment, a computer device is provided, which may be a terminal, and the internal structure of which may be as shown in fig. 3. The computer device includes a processor, a memory, a communication interface, a display screen, and an input device connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The communication interface of the computer device is used for communicating with an external terminal in a wired or wireless manner, which may be implemented by WI F I, mobile cellular network, NFC (near field communication) or other technologies. The computer program is executed by a processor to implement a road feature acquisition method. The display screen of the computer equipment can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer equipment can be a touch layer covered on the display screen, can also be keys, a track ball or a touch pad arranged on the shell of the computer equipment, and can also be an external keyboard, a touch pad or a mouse and the like.
It will be appreciated by those skilled in the art that the structure shown in FIG. 3 is merely a block diagram of some of the structures associated with the present inventive arrangements and is not limiting of the computer device to which the present inventive arrangements may be applied, and that a particular computer device may include more or fewer components than shown, or may combine some of the components, or have a different arrangement of components.
In an embodiment, a computer device is provided comprising a memory and a processor, the memory having stored therein a computer program, the processor implementing the steps of the method of any of the first aspects when the computer program is executed.
In an embodiment, a computer-readable storage medium is provided, on which a computer program is stored which, when executed by a processor, implements the steps of the method of any of the first aspects.
In an embodiment, a computer program product is provided, comprising a computer program which, when executed by a processor, implements the steps of the method of any of the first aspects.
The user information (including but not limited to user equipment information, user personal information, etc.) and the data (including but not limited to data for analysis, stored data, presented data, etc.) related to the present application are information and data authorized by the user or sufficiently authorized by each party.
Those skilled in the art will appreciate that implementing all or part of the above described methods may be accomplished by way of a computer program stored on a non-transitory computer readable storage medium, which when executed, may comprise the steps of the embodiments of the methods described above. Any reference to memory, database, or other medium used in embodiments provided herein may include at least one of non-volatile and volatile memory. The nonvolatile Memory may include Read-On-i-y Memory (ROM), magnetic tape, floppy disk, flash Memory, optical Memory, high density embedded nonvolatile Memory, resistive random access Memory (ReRAM), magneto-resistive random access Memory (Magnetores i st i ve Random Access Memory, MRAM), ferroelectric Memory (Ferroe l ectr i c Random Access Memory, FRAM), phase change Memory (PHASE CHANGE Memory, PCM), graphene Memory, and the like. Volatile memory can include random access memory (Random Access Memory, RAM) or external cache memory, and the like. By way of illustration, and not limitation, RAM can be in various forms such as static random access memory (Stat i c Random Access Memory, SRAM) or dynamic random access memory (Dynami c Random Access Memory, DRAM), etc. The databases referred to in the embodiments provided herein may include at least one of a relational database and a non-relational database. The non-relational database may include, but is not limited to, a blockchain-based distributed database, and the like. The processor referred to in the embodiments provided in the present application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic unit, a data processing logic unit based on quantum computing, or the like, but is not limited thereto.
The technical features of the above embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.
The foregoing examples illustrate only a few embodiments of the application and are described in detail herein without thereby limiting the scope of the application. It should be noted that it will be apparent to those skilled in the art that several variations and modifications can be made without departing from the spirit of the application, which are all within the scope of the application. Accordingly, the scope of the application should be assessed as that of the appended claims.
Claims (8)
1. A method of obtaining road characteristics, the method comprising:
acquiring image information of a target road, spatial structure information of the target road and radiation wave detection information of the target road;
identifying each defect characteristic image of the image information and an image range of each defect characteristic image through a self-attention mechanism, and identifying each abnormal three-dimensional fluctuation range of the target road based on the radiowave detection information and the spatial structure information of the target road;
Determining each target damage area of the target road in the spatial structure information based on each abnormal three-dimensional fluctuation range and each image range of the defect characteristic image, and determining damage information corresponding to each target damage area based on the image information of the target road;
extracting characteristic information of each target damage area based on the damage information of each target damage area, and taking the characteristic information of all the target damage areas as target characteristic information of the target road;
the identifying, based on the radiation wave detection information and the spatial structure information of the target road, the respective abnormal three-dimensional fluctuation ranges of the target road includes:
Projecting the radiation wave detection information into the space structure information to obtain a three-dimensional structure fluctuation image of the target road;
Identifying each abnormal structure fluctuation image in the three-dimensional structure fluctuation image, and identifying the image edge of each abnormal structure fluctuation image through an image edge algorithm;
taking the range included by the image edge of each abnormal structure fluctuation image as an abnormal three-dimensional fluctuation range;
The determining, in the spatial structure information, each target damaged area of the target road based on each of the abnormal three-dimensional fluctuation ranges and the image range of each of the defect feature images, includes:
performing dimension reduction processing on the three-dimensional structure fluctuation image to obtain a two-dimensional structure image of the three-dimensional structure fluctuation image, fusing the two-dimensional structure image with the image information, and judging whether a sub-two-dimensional structure image corresponding to an abnormal three-dimensional fluctuation range intersected with the defect characteristic image exists or not;
Under the condition that a sub-two-dimensional structural image corresponding to an abnormal three-dimensional fluctuation range intersected with a defect characteristic image exists, carrying out image range adjustment processing on the two-dimensional image range corresponding to the sub-two-dimensional structural image intersected with the defect characteristic image based on the image range of each defect characteristic image to obtain each two-dimensional target damage area, and projecting each two-dimensional target damage area into the three-dimensional structural fluctuation image to obtain each target damage area of the target road;
under the condition that sub two-dimensional structure images corresponding to abnormal three-dimensional fluctuation ranges intersecting with the defect feature images do not exist, projecting the image ranges of the defect feature images into the space structure information, identifying defect feature three-dimensional structure information corresponding to each defect feature image, and taking the three-dimensional image ranges corresponding to the defect feature three-dimensional structure information as a first target damage area of the target road;
And regarding the three-dimensional image range of each abnormal three-dimensional fluctuation range in the space structure information as a second target damage area, and regarding all first target damage areas and all second target damage areas as each target damage area of the target road.
2. The method of claim 1, wherein identifying each defect feature image of the image information and an image range of each defect feature image by a self-attention mechanism comprises:
inputting the image information into a self-attention network to obtain each initial defect image in the image information, and identifying the image characteristics of each initial defect image through an image characteristic identification network;
Taking initial defect images corresponding to image features meeting defect feature information as defect images, and based on the position information of each defect image in the image information of the target road, performing splicing processing on the defect images connected with the position information to obtain each defect feature image;
and identifying the image range of each defect characteristic image based on the image information.
3. The method according to claim 1, wherein determining damage information corresponding to each target damage region based on the image information of the target road includes:
Performing dimension reduction treatment on each target damage region to obtain two-dimensional target damage regions corresponding to each target damage region, and inquiring damage ranges corresponding to each two-dimensional target damage region in the image information;
and identifying the damage type of each damage range through an image identification strategy, and taking the damage range containing the damage type as damage information corresponding to each target damage area.
4. The method of claim 3, wherein extracting feature information of each of the target lesion areas based on the lesion information of each of the target lesion areas comprises:
Extracting image features of each target damage region through a feature recognition strategy, and adding feature labels to the image features based on the damage range of each target damage region and the damage type of each damage range to obtain feature information of each target damage region.
5. A road feature acquisition apparatus, the apparatus comprising:
The acquisition module is used for acquiring image information of a target road, spatial structure information of the target road and radiation wave detection information of the target road;
The identification module is used for identifying each defect characteristic image of the image information and the image range of each defect characteristic image through a self-attention mechanism, and identifying each abnormal three-dimensional fluctuation range of the target road based on the radiation wave detection information and the spatial structure information of the target road;
The determining module is used for determining each target damage area of the target road in the space structure information based on each abnormal three-dimensional fluctuation range and the image range of each defect characteristic image, and determining damage information corresponding to each target damage area based on the image information of the target road;
the extraction module is used for extracting the characteristic information of each target damage area based on the damage information of each target damage area, and taking the characteristic information of all the target damage areas as the target characteristic information of the target road;
The identification module is used for identifying the specific implementation of each abnormal three-dimensional fluctuation range of the target road based on the radiation wave detection information and the spatial structure information of the target road, wherein the specific implementation is as follows:
Projecting the radiation wave detection information into the space structure information to obtain a three-dimensional structure fluctuation image of the target road;
Identifying each abnormal structure fluctuation image in the three-dimensional structure fluctuation image, and identifying the image edge of each abnormal structure fluctuation image through an image edge algorithm;
taking the range included by the image edge of each abnormal structure fluctuation image as an abnormal three-dimensional fluctuation range;
the determining module determines, in the spatial structure information, specific implementation of each target damage area of the target road based on each abnormal three-dimensional fluctuation range and an image range of each defect feature image, as follows:
performing dimension reduction processing on the three-dimensional structure fluctuation image to obtain a two-dimensional structure image of the three-dimensional structure fluctuation image, fusing the two-dimensional structure image with the image information, and judging whether a sub-two-dimensional structure image corresponding to an abnormal three-dimensional fluctuation range intersected with the defect characteristic image exists or not;
Under the condition that a sub-two-dimensional structural image corresponding to an abnormal three-dimensional fluctuation range intersected with a defect characteristic image exists, carrying out image range adjustment processing on the two-dimensional image range corresponding to the sub-two-dimensional structural image intersected with the defect characteristic image based on the image range of each defect characteristic image to obtain each two-dimensional target damage area, and projecting each two-dimensional target damage area into the three-dimensional structural fluctuation image to obtain each target damage area of the target road;
under the condition that sub two-dimensional structure images corresponding to abnormal three-dimensional fluctuation ranges intersecting with the defect feature images do not exist, projecting the image ranges of the defect feature images into the space structure information, identifying defect feature three-dimensional structure information corresponding to each defect feature image, and taking the three-dimensional image ranges corresponding to the defect feature three-dimensional structure information as a first target damage area of the target road;
And regarding the three-dimensional image range of each abnormal three-dimensional fluctuation range in the space structure information as a second target damage area, and regarding all first target damage areas and all second target damage areas as each target damage area of the target road.
6. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor implements the steps of the method of any of claims 1 to 4 when the computer program is executed.
7. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1 to 4.
8. A computer program product comprising a computer program, characterized in that the computer program, when executed by a processor, implements the steps of the method of any of claims 1 to 4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310900214.3A CN117197774B (en) | 2023-07-19 | 2023-07-19 | Road feature acquisition method and device, computer equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310900214.3A CN117197774B (en) | 2023-07-19 | 2023-07-19 | Road feature acquisition method and device, computer equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117197774A CN117197774A (en) | 2023-12-08 |
| CN117197774B true CN117197774B (en) | 2024-10-18 |
Family
ID=89004120
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310900214.3A Active CN117197774B (en) | 2023-07-19 | 2023-07-19 | Road feature acquisition method and device, computer equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117197774B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118781772B (en) * | 2024-09-11 | 2024-12-10 | 辽宁安舟数字科技有限公司 | Port intelligent safety supervision and emergency response system and method |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115063525A (en) * | 2022-04-06 | 2022-09-16 | 广州易探科技有限公司 | Three-dimensional mapping method and device for urban road subgrade and pipeline |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2013090830A1 (en) * | 2011-12-16 | 2013-06-20 | University Of Southern California | Autonomous pavement condition assessment |
| CN108572248A (en) * | 2018-04-16 | 2018-09-25 | 长沙理工大学 | A Method for Evaluating Technical Condition of Roads Based on Nondestructive Testing Technology |
| CN113780178B (en) * | 2021-09-10 | 2024-10-29 | 北京百度网讯科技有限公司 | Road detection method, device, electronic equipment and storage medium |
| CN115097445B (en) * | 2022-06-20 | 2024-07-16 | 中国铁建港航局集团有限公司 | Three-dimensional ground penetrating radar detection method, system, equipment and terminal for road subgrade diseases |
-
2023
- 2023-07-19 CN CN202310900214.3A patent/CN117197774B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115063525A (en) * | 2022-04-06 | 2022-09-16 | 广州易探科技有限公司 | Three-dimensional mapping method and device for urban road subgrade and pipeline |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117197774A (en) | 2023-12-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Ghosh Mondal et al. | Deep learning‐based multi‐class damage detection for autonomous post‐disaster reconnaissance | |
| Huang et al. | Road centreline extraction from high‐resolution imagery based on multiscale structural features and support vector machines | |
| Su et al. | Deep convolutional neural network–based pixel-wise landslide inventory mapping | |
| Peraka et al. | Development of a multi-distress detection system for asphalt pavements: Transfer learning-based approach | |
| CN117197774B (en) | Road feature acquisition method and device, computer equipment and storage medium | |
| Chen et al. | Extraction of bridges over water from high-resolution optical remote-sensing images based on mathematical morphology | |
| Doycheva et al. | Implementing textural features on GPUs for improved real-time pavement distress detection | |
| CN115984273B (en) | Road disease detection method, device, computer equipment and readable storage medium | |
| Ejimuda et al. | Using deep learning and computer vision techniques to improve facility corrosion risk management systems 2.0 | |
| Ashraf et al. | Machine learning-based pavement crack detection, classification, and characterization: a review | |
| CN114898357B (en) | Defect identification method and device, electronic equipment and computer readable storage medium | |
| Wang et al. | Image analysis of the automatic welding defects detection based on deep learning | |
| Espindola et al. | Comparing different deep learning architectures as vision-based multi-label classifiers for identification of multiple distresses on asphalt pavement | |
| Li et al. | A geometrical morphology-enhanced computer vision approach for structural health assessment | |
| CN115423798A (en) | Defect identification method, defect identification device, computer equipment, storage medium and computer program product | |
| Maurya et al. | A global context and pyramidal scale guided convolutional neural network for pavement crack detection | |
| Qiao et al. | Revolutionizing building damage detection: A novel weakly supervised approach using high-resolution remote sensing images | |
| CN113792169A (en) | Digital archive management method and system based on big data application | |
| Fahmani et al. | Deep learning-based predictive models for pavement patching and manholes evaluation | |
| JP6434834B2 (en) | Inspection object extraction device and inspection object extraction method | |
| CN116610583A (en) | SCA tool maturity evaluation method, SCA tool maturity evaluation device, SCA tool maturity evaluation equipment, SCA tool maturity evaluation medium and SCA tool maturity evaluation product | |
| Sekar et al. | A novel SGD-U-network-based pixel-level road crack segmentation and classification | |
| CN114630102A (en) | Method and device for detecting angle change of data acquisition equipment and computer equipment | |
| CN118052816B (en) | PCBA surface defect detection method and system | |
| CN115965856B (en) | Image detection model construction method, device, computer equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |