CN109870456B - Rapid detection system and method for road surface health condition - Google Patents
Rapid detection system and method for road surface health condition Download PDFInfo
- Publication number
- CN109870456B CN109870456B CN201910105679.3A CN201910105679A CN109870456B CN 109870456 B CN109870456 B CN 109870456B CN 201910105679 A CN201910105679 A CN 201910105679A CN 109870456 B CN109870456 B CN 109870456B
- Authority
- CN
- China
- Prior art keywords
- vehicle
- road surface
- data
- road
- pavement
- 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
Images
Abstract
The invention discloses a system and a method for rapidly detecting the health condition of a road surface, which comprises an acquisition module: the method is used for acquiring vibration acceleration, pavement damage videos and GPS positioning data in the vehicle; the vehicle-mounted terminal: the cloud server is used for receiving the data of each acquisition module, preprocessing the acquired data and uploading the processed data to the cloud server; a server on the cloud: the system is used for receiving, analyzing and storing data uploaded by the vehicle-mounted terminal; and a visualization terminal: the system is used for publishing data and generating reports, and providing real-time, multi-dimensional and customized display detection results; the acquisition module is connected with the vehicle-mounted terminal through the transmission module, and the vehicle-mounted terminal, the cloud server and the visual terminal are connected through a network. The invention integrates data acquisition, transmission, analysis and release, greatly reduces the detection cost, and can conveniently provide accurate and high-frequency road pavement condition and road safety information for the social public.
Description
Technical Field
The invention relates to a pavement detection system and a method, in particular to a pavement health condition rapid detection system and a method.
Background
In recent years, maintenance decisions based on data driving have become national policy requirements, but China still faces huge detection pressure, only about 10% of expressways and high-grade roads are effectively detected in a road network of more than 450 kilometers, but a proper detection means is lacked for large-scale low-grade roads and village and town roads, and only a few big cities carry out daily detection and maintenance on urban roads.
At present, the price of mainstream road surface detection methods such as laser detectors, three-dimensional radar detectors and the like is usually different from 100 + 1200 ten thousand, the price of multifunctional detection vehicles developed by the Australian ARRB group is up to 2400 ten thousand, and the mainstream road surface detection methods are difficult to introduce in large quantities and are used for rapid detection of road quality. On the other hand, many conventional detection methods, such as a three-meter ruler, a level gauge, a hand-push type profiler and the like, are low in price, but due to the semi-manual and semi-mechanical operation mode, the detection efficiency is low, and the detection method is generally only used for calibrating equipment and cannot be popularized in a large range.
In addition, the domestic road management system has the defects of high cost and low efficiency. The road detection data has low data interaction degree due to the difference of regions, and excellent maintenance and repair decisions cannot be widely applied. In recent years, with the rapid rise of the concept of "internet +", the national and local road administration has gradually opened up much road administration data. The data storage is huge, and if reasonable mining and utilization can be carried out, huge convenience and benefits can be brought to the society.
Disclosure of Invention
The invention aims to solve the technical problem of providing a system and a method for rapidly detecting the road health condition, which integrate data acquisition, transmission, analysis and release, greatly reduce the detection cost and conveniently provide accurate and high-frequency road condition and road safety information for social public.
The invention adopts the technical scheme to solve the technical problems and provides a pavement health condition rapid detection system, which comprises an acquisition module: the method is used for acquiring vibration acceleration, pavement damage videos and GPS positioning data in the vehicle; the vehicle-mounted terminal: the cloud server is used for receiving the data of each acquisition module, preprocessing the acquired data and uploading the processed data to the cloud server; a server on the cloud: the system is used for receiving, analyzing and storing data uploaded by the vehicle-mounted terminal; and a visualization terminal: the system is used for publishing data and generating reports, and providing real-time, multi-dimensional and customized display detection results; the acquisition module is connected with the vehicle-mounted terminal through the transmission module, and the vehicle-mounted terminal, the cloud server and the visual terminal are connected through a network.
Furthermore, the acquisition module comprises an in-vehicle vibration acceleration acquisition module, a pavement damage video acquisition module and a GPS positioning acquisition module.
Further, the in-vehicle vibration acceleration acquisition module is two triaxial MEMS vibration sensors that are symmetrically distributed, and two triaxial MEMS vibration sensors are fixed on the vehicle trunk and are located above the rear wheels, so that the x axis of the vibration sensor is parallel to the direction of the vehicle body, and the z axis is parallel to the direction of the vehicle height.
Furthermore, the pavement disease video acquisition module is an industrial digital camera, the resolution of the industrial digital camera is more than or equal to 200 ten thousand pixels, the maximum pixel is less than or equal to 7.5 mu m, and the aperture is more than or equal to F1.4.
The invention also provides a method for rapidly detecting the health condition of the road surface to solve the technical problems, wherein the in-vehicle vibration acceleration acquisition module is two three-axis MEMS vibration sensors which are symmetrically distributed, the road surface disease video acquisition module is an industrial digital camera, and the vehicle-mounted terminal stores acquired and processed data into a local GPS database, a flatness database and a road surface condition index database; the detection method comprises the following steps: s1) road surface evenness collection: the method comprises the following steps that two MEMS vibration sensors acquire vibration information of a vehicle, and a vehicle-mounted terminal calculates an international road surface flatness index IRI by using a power spectral density algorithm; s2) road surface disease condition collection: collecting a pavement picture by a vehicle-mounted industrial digital camera, identifying pavement diseases in the picture through a depth neural network algorithm, and calculating a pavement condition index PCI according to the pavement diseases; s3) data uploading: in the detection process, the vehicle-mounted terminal continuously accesses a local GPS database, a flatness database and a road surface condition index database, and when records which are not uploaded exist in the databases, the vehicle-mounted terminal calls out the data and uploads the data to the cloud-end platform through the transmission module; s4) visually displaying: the cloud server matches the road surface condition index PCI, the road surface smoothness data and the GPS data, converts the international smoothness index IRI into a road surface running quality index RQI and matches the road surface running quality index RQI with the GPS data, and further matches the road surface condition index PCI and the road surface running quality index RQI with an electronic road section in a GIS map, so that the road surface smoothness and the road surface damage condition can be visually displayed.
Further, the step S1 includes: s11) the GPS positioning acquisition module continuously acquires the position information of the vehicle and continuously calculates the assumed accumulated displacement of the vehicle; starting the vehicle from rest, and accumulating the displacement from 0; s12) continuously acquiring vibration data of the vehicle by the two vehicle-mounted MEMS vibration sensors, and transmitting the data to the vehicle-mounted terminal; the acquisition frequency of the flatness data can be adjusted according to the vehicle speed: when the vehicle speed is 0, the acquisition frequency is 0 Hz; when the vehicle speed is 3.6km/h, the acquisition frequency is 20 Hz; when the vehicle speed is 36km/h, the acquisition frequency is 200 Hz; when the vehicle speed is v km/h, the acquisition frequency is
S13) when the accumulated displacement reaches or exceeds the preset distance, the vehicle-mounted terminal stores the vibration data collected all the time into a section of data to be processed, processes the data by using a power spectral density algorithm, and calculates the international flatness index IRI of the sectioniMeanwhile, the centroid position of the road section is calculated according to the position information collected by the positioning sensor, and the flatness calculation result and the corresponding road section centroid position are stored in a local flatness database; s14), the accumulated displacement is reset to zero, and the steps S12 and S13 are repeated to start the position acquisition of the next link.
Further, step S1 further includes performing a calibration test for the detected vehicle periodically, and adjusting the relevant parameters of the power spectrum algorithm, where the calibration test is performed by: international flatness index IRI for a vehicle at a certain length0When the vehicle runs on a known road section, the vibration sensor collects the vibration information of the vehicle and calculates IRI through a power spectral density algorithm1And with IRI0And (3) comparison:if the result satisfies
The vehicle vibration parameter calibration result is considered to be accurate; otherwise, the vibration parameters of the vehicle are continuously adjusted until the vibration parameters are met
When calibration collection is carried out, the vehicle speeds are respectively controlled to be 20km/h, 40km/h and 60km/h, the vehicle runs at a constant speed, and at least 2 times of collection are carried out under each vehicle speed condition.
Further, the step S2 includes: s21) the positioning acquisition sensor continuously acquires the position information of the vehicle and continuously calculates the assumed accumulated displacement of the vehicle; starting the vehicle from rest, and accumulating the displacement from 0; s22) the industrial digital camera collects road surface pictures, and the frequency of collecting the road surface pictures can be adjusted according to the vehicle speed; when the vehicle speed is 0, the image acquisition frequency is 0; when the vehicle speed is 3.6km/h, the picture acquisition frequency is 1 Hz; when the vehicle speed is v km/h, the acquisition frequency of the pictures is (20 x v)/3.6 Hz; s23), in the detection process, the collected image is stored in a local folder, the position information collected by the positioning sensor is stored in a local GPS database, and the collected image path corresponding to the GPS is written into the GPS database; s24) when the accumulated displacement reaches or exceeds a preset distance, the vehicle-mounted terminal inputs all the images collected in the road section into a neural network algorithm, identifies the disease of the ith type of road surface through the neural network algorithm and calculates the area A of the disease areaiCorresponding degree of disease wiCalculating the road surface damage rate DR of the road sectioniAnd calculating the road surface condition index PCI of the road sectioniMeanwhile, the centroid position of the road section is calculated according to the position information acquired by the positioning sensor; the calculation result of the road surface condition index PCI and the corresponding road section centroid position are stored in a local road surface condition index database.
Further, the road picture in step S3 is collected as follows: s31) according to the transformation relation of the photo visual angle and distortion to the photo in the camera internal reference calibration process, carrying out visual angle transformation processing on the photoThe effective area after the visual angle of the photo is changed is A; s32) identifying the disease type in the effective area of the picture after the visual angle transformation by using the trained neural network, and identifying the type A of the ith type of pavement disease in the pictureiCorresponding disease degree wiAnd calculating the road surface damage rate according to the following formula:
in the formula, DR is the pavement damage rate and is the percentage of the sum of the whole damage area caused by various damages and the effective area in the picture; w is aiWeighting the damage of the i-th road surface; s33) the vehicle-mounted terminal calculates the PCI of the road segment according to the accumulated displacement, specifically: assuming that the vehicle starts from a standstill, the displacement is accumulated from 0, when the accumulated displacement reaches or exceeds 20m, the road surface damage condition index PCI in the accumulated journey is calculated, the calculation result and the corresponding GPS data are recorded in a local PCI database, and meanwhile, the accumulated displacement of the vehicle-mounted terminal is accumulated again from 0; the PCI calculation formula is as follows: PCI 100-a0DRa1(ii) a Wherein DR is a road surface breakage rate, a0、a1Respectively road surface type coefficients; s34) the accumulated displacement is reset to zero, and the step S33) is repeated, and calculation of the road surface condition index of the next road section is started.
Further, the step S3 further includes performing the following correction on the view angle and distortion of the captured picture: s301) placing checkerboard calibration cloth at a proper position in the middle of a picture and shooting the picture; s302) searching the angular points of the checkerboard, and searching four angular points at the outermost edge; s303) automatically finding and only considering the vertical straight line of the rectangular area surrounded by the four corner points; s304) according to the projection principle of camera shooting, each vertical straight line is finally converged at one point to find the point; s305) respectively connecting a convergence point with two angular points of the upper edge of the image, wherein the extension line of each connecting line is intersected with the extension line of the lower edge of the image at a point, and the two angular points of the upper edge of the image and the two intersection points on the extension line of the lower edge form a trapezoid which comprises the whole picture and represents the projection angle of the whole picture; s306) converting the projection visual angle into a overlooking visual angle, converting two inclined edges of the trapezoid into two parallel lines, and controlling the distance between an upper parallel edge and a lower parallel edge of the trapezoid to enable the proportion of the horizontal edge and the vertical edge of the checkerboard to still keep the actual proportion; s307) searching angular points of the checkerboard after the angle of view is converted, calculating the transverse pixel distance between the two angular points and the number of the checkerboard between the two angular points, thereby obtaining the pixel number corresponding to the transverse distance/vertical distance of one checkerboard in a top view; corresponding to the actual size of the checkerboard, and acquiring the actual size corresponding to the horizontal pixel distance/vertical pixel distance in the top view; s308) acquiring the actual size or area corresponding to the road surface damage in the picture shot at the visual angle according to the corresponding relation between the pixels acquired in the step S307) and the actual size of the calibration checkerboard, and further acquiring the road surface area corresponding to the whole shot picture.
Compared with the prior art, the invention has the following beneficial effects: the invention provides a rapid detection system and a rapid detection method for road surface health conditions, which adopt a lightweight framework mode and integrate data acquisition, transmission, analysis and release into a whole; by the aid of technologies such as machine vision, deep learning and cloud computing, rapid detection of key pavement health indexes such as pavement evenness, bridge head bumping, pavement apparent diseases and pavement structure depth can be achieved, and detection and analysis results are displayed in real time through a cloud platform. The equipment price of the lightweight detection system is only 10% -30% of the price of the similar equipment in the market, so that the detection cost is greatly reduced. The detection data can be fused with the multi-element data, the detection result can be checked at the release terminal and the information platform in real time, accurate and high-frequency road pavement condition and road safety information are provided for the social public, a scientific maintenance plan is made for road maintenance enterprises, and data support is provided for government financial decisions.
Drawings
FIG. 1 is a block diagram of a rapid road health status detection system according to the present invention;
FIG. 2 is a photograph of the present invention after correcting using internal references;
FIG. 3 is a schematic diagram of the present invention looking for a vertical line;
FIG. 4 is a top view of the checkerboard of the present invention;
FIG. 5 is a schematic diagram of the present invention for transforming the view angle and distortion of the collected picture.
Detailed Description
The invention is further described below with reference to the figures and examples.
FIG. 1 is a block diagram of a rapid road health status detection system according to the present invention.
Referring to fig. 1, the system for rapidly detecting the health status of the road surface provided by the invention comprises an acquisition module, a transmission module, a vehicle-mounted terminal, a cloud server and a visual terminal, wherein the acquisition module further comprises: the method comprises the steps of vibration acceleration acquisition in a vehicle, pavement damage video acquisition and GPS positioning acquisition.
The pavement health condition rapid detection system provided by the invention has the advantages that the detection error is +/-10%, the requirement on the specified precision of road maintenance can be met, the detection efficiency is high, the measurement of more than 300 kilometers can be realized on a single vehicle per day, and besides vehicle-mounted power supply, the sensor module is powered by the storage battery and can continuously and stably work for more than 72 hours. In addition, the invention has the characteristics of low cost, light weight, customization as required and one-key operation, and can realize the full-coverage detection of the road surface state of the road network and the informatization of the road traffic infrastructure condition; meanwhile, detection results can be fused and mined with multi-source data, the detection results can be checked at a release terminal and an information platform in real time, customized services are conveniently provided for various large road management units and detection institutions, road pavement conditions and road safety information are provided for social public, and data support is provided for government financial decisions.
The invention provides a pavement health condition rapid detection system, which comprises the following main modules:
1 acquisition Module
1.1 in-vehicle vibration acceleration acquisition
And measuring the three-axis acceleration of the vehicle in the running process through the vehicle-mounted portable data acquisition device. The related technical requirements of the portable data collector are as follows:
(1) the portable data acquisition unit can measure the three-axis acceleration, and the measuring range is +/-5.0 g (g is the gravity acceleration, and the value is 9.8 m.s)-2)；
(2) The acceleration measurement precision is less than or equal to 0.1 g;
(3) the acceleration measurement frequency is 0-1000 Hz adjustable;
(4) the collector supplies power through a serial port;
(5) the collector can be modified according to different vehicle specifications, and an intelligent MEMS module is embedded in the collector;
(6) the collectors are generally arranged above the appointed wheel shafts of the vehicle, the number of the collectors is two or more, and the collectors are fixedly connected with the vehicle body;
(7) the collector transmits the collected data to the vehicle-mounted terminal in a wired or wireless mode.
1.2 pavement disease video acquisition
And shooting and collecting road video information through a high-speed industrial camera. The related technical requirements of the high-speed industrial camera are as follows:
(1) the resolution of the used industrial camera is more than or equal to 200 ten thousand pixels, the maximum pixel is less than or equal to 7.5 mu m, and the aperture is more than or equal to F1.4;
(2) the shooting frame rate of the camera is 1-100 frames per second-1The internal adjustment is realized, the actual shooting frame rate can be adjusted according to the vehicle speed, and the maximum vehicle speed of 80 km.h is met-1Continuous shooting on the road surface is realized without missing shooting;
(3) the used camera has editability and sufficient I/O interfaces, and secondary development is allowed;
(4) the camera has two photo collecting modes, one mode is that the camera is electrified and then continuously collects photos, the other mode is that the camera is triggered by the vehicle speed to collect photos, and the collecting frequency changes along with the vehicle speed.
(5) The camera can transmit and store in real time and transmit the data to the vehicle-mounted terminal in a wired or wireless manner; the method supports triggering image storage and video recording, and meanwhile supports breakpoint transmission by self-contained cache.
(6) The cameras are generally mounted on the top of the vehicle with the lens facing the ground behind and below the vehicle, and the number of the cameras is one or more.
1.3 positioning acquisition module
The positioning acquisition module can acquire the position information of the vehicle in real time. The technical requirements are as follows: (1) GPS and Beidou dual-mode positioning is adopted; (2) the receiver is positioned by combining inertial navigation and difference, and the positioning error is less than or equal to 1.0 m; (3) the data acquisition frequency can be adjusted to be 1Hz or 5 Hz; (4) the receivers are typically mounted on a test cart, typically one in number.
2 transmission module
The transmission module is used for realizing data and information transmission among different components and different modules. The transmission module used by the system comprises wired transmission and wireless transmission. The technical requirements are as follows: (1) wired transmission: the transmission efficiency is more than or equal to gigabit Ethernet (GigE) transmission, and the signal-to-noise ratio is high; (2) wireless transmission: and packaging the data to be transmitted, and transmitting the data through the 3/4/5G network.
3 vehicle terminal
The vehicle-mounted terminal is an industrial host and is used for receiving the data acquired by each acquisition module and acquiring the state information of the test system; cleaning, processing and analyzing the acquired data; and uploading the analyzed data to a cloud server. The relevant technical requirements of the vehicle-mounted terminal are as follows: (1) the industrial host computer comprises a Graphic Processing Unit (GPU) and a plurality of data interfaces, and is provided with a wire and wireless transmission interface; (2) the system has a local storage function, and can store data in a network-free environment; (3) the vehicle-mounted terminal is generally arranged on a front seat of a vehicle, and the number of the vehicle-mounted terminal is generally one; (4) the vehicle-mounted terminal is powered by a cigarette lighter on the test vehicle and supports hot plugging.
4 server on cloud
The cloud server has the function of receiving, analyzing and storing data uploaded by the vehicle-mounted terminal. The technical requirements of the server on the cloud are as follows: (1) the uploading data can be received in real time; (2) carrying out encryption processing on the database; (3) the data of the database can be accessed at various network terminals; (4) the cloud upper platform is provided with a remote disaster backup data synchronization and platform double-active undisturbed switching mechanism, namely bottom layer monitoring data is synchronously stored in a main center storage system and a disaster backup center storage system, the continuity and integrity of data of a main monitoring center and a backup monitoring center are ensured, the data loss probability is reduced as much as possible, the double centers are synchronously copied, and the data are synchronized in real time; under normal conditions, the client accesses the application server of the main center through the ring network, and when a disaster occurs, the client can directly access the application server of the disaster recovery center through network switching; (5) and the cloud platform can interact with other multivariate data, and is beneficial to further mining of the data.
5 visual terminal
The web development visual query end is utilized, detection results can be displayed on various terminal platforms such as a computer, a mobile phone and a tablet in real time, in multiple dimensions, in deep level and in a customized mode, road pavement conditions and road safety information are provided for social public, scientific maintenance plans are made for road maintenance enterprises, and data support is provided for government financial decisions.
The method is based on rapid pavement evenness detection of a distributed sensing network, and the pavement evenness condition is inversely calculated by collecting multi-point in-vehicle vibration acceleration data and combining a spectral density analysis algorithm; and recording the abnormal vibration condition of the vehicle, and locking the bumping site and degree of the bridge head by combining the GPS layer information. Different from the traditional laser detection method, the measurement error of the method is kept within +/-10%, the measurement efficiency exceeds 300km/day, and the efficiency and the sustainability of the pavement evenness detection are greatly improved.
The invention effectively solves the problems of time and labor consumption, complex operation, high price and the like of the traditional equipment, can realize detection with multiple vehicles, large range, time and low consumption, fills the blank of the flatness detection means of roads of low grade and roads of villages and towns in China, and can effectively improve the technical level of road whole-life management in China.
The invention discloses a distributed sensing network-based rapid pavement evenness detection method, which is mainly completed by the cooperation of four sub-modules:
(1) the portable data acquisition unit: acquiring three-axis vibration acceleration in the vehicle;
(2) AI vehicle terminal: preprocessing data, matching the data with GPS data, and uploading the data to a cloud;
(3) a server on the cloud: receiving, analyzing and storing data;
(4) visual terminal: and (5) data publishing and report generation.
The invention provides a pavement health condition rapid detection system, which comprises the following control processes:
1. installation of equipment
Two intelligent three-axis MEMS vibration sensors are mounted above a rear wheel of a test vehicle trunk, so that the x axis of each vibration sensor is parallel to the direction of a vehicle body, the z axis of each vibration sensor is parallel to the direction of the vehicle height, and the two sensors are symmetrically distributed. The vehicle-mounted terminal, the transmission module and the positioning module are also installed in place.
2. Vibration parameter calibration
In order to ensure that the calculation result of the road surface evenness is effective and reliable, calibration test needs to be carried out on a detected vehicle regularly, and relevant parameters of a power spectrum algorithm are adjusted. The calibration test method comprises the following steps: international flatness index IRI for a vehicle at a certain length0The method comprises the steps that a test vehicle runs on a calibrated test road, a system is opened in the running process, vehicle body vibration data are collected through two intelligent three-axis MEMS vibration sensors, and speed and displacement data of the vehicle are collected through a GPS. Segmenting the road to be measured every 100m, and trial calculating the flatness IRI of the road surface in each segmented road by using a power spectrum density algorithm according to the speed, the displacement, the vibration data and the preliminary vehicle vibration parametersiAnd the calibrated road flatness IRI0Comparing, and when the trial calculation result in each road section meets the requirement
The vehicle vibration parameter calibration result is considered to be accurate; otherwise, the vibration parameters of the vehicle are continuously adjusted until the vibration parameters are met
In order to ensure a calibration result, when calibration acquisition is carried out, the speed is respectively controlled at 20km/h, 40km/h and 60km/h, the vehicle runs at a constant speed, and at least 2 times of acquisition is carried out under each speed condition.
3. Vibration measurement and flatness calculation
When the test is formally started, the triaxial MEMS vibration sensor transmits the acquired triaxial vibration data of the vehicle to the vehicle-mounted terminal, and the positioning module transmits the real-time position data and speed data of the vehicle to the vehicle-mounted terminal. And the vehicle-mounted terminal calculates the international flatness index IRI of the road section by using a power spectral density algorithm according to the vehicle vibration parameter, the vehicle vibration data, the vehicle position data and the speed data. The method comprises the following specific steps:
(1) the positioning acquisition sensor continuously acquires the position information of the vehicle and continuously calculates the assumed accumulated displacement of the vehicle; starting the vehicle from a standstill, and accumulating the displacement from 0;
(2) the two vehicle-mounted MEMS vibration sensors continuously acquire vibration data of the vehicle and transmit the data to the vehicle-mounted terminal; the acquisition frequency of the flatness data can be adjusted according to the vehicle speed: when the vehicle speed is 0, the acquisition frequency is 0 Hz; when the vehicle speed is 3.6km/h, the acquisition frequency is 20 Hz; when the vehicle speed is 36km/h, the acquisition frequency is 200 Hz; when the vehicle speed is v km/h, the acquisition frequency is
(3) When the accumulated displacement reaches or exceeds a preset distance, the vehicle-mounted terminal stores the vibration data which are collected all the time into a section of data to be processed, the data are processed by using a power spectral density algorithm, and the international flatness index IRI of the road section is calculatediAnd meanwhile, the centroid position of the road section is calculated according to the position information acquired by the positioning sensor. The flatness calculation result and the position of the road section centroid corresponding to the flatness calculation result are stored in a local flatness database.
(4) And (5) resetting the accumulated displacement to zero, repeating the step (2) and starting the position acquisition of the next road section.
4. Data matching and uploading
And writing the data in the flatness database into a local SD card or hard disk of the vehicle-mounted terminal, and simultaneously uploading the detected data to a cloud server in real time through a 4G network of the vehicle-mounted terminal.
5. Data summarization and visual display
After receiving a connection request of a remote vehicle-mounted terminal, the on-cloud server verifies the terminal identity, and if the terminal is a legal terminal, a data receiving thread is established for the terminal to receive data; and if the terminal is an illegal terminal, closing the link. The received international flatness index IRI data is converted into a road surface running quality index RQI, the road surface running quality index RQI is led into a GIS layer and matched with a corresponding electronic road section, and the distribution condition of the road network running quality index RQI is visually displayed through a system platform.
The power spectral density algorithm is used creatively in the rapid detection of the road surface evenness, the vehicle can run at variable speed in the detection process, the vibration data can rapidly calculate the road surface evenness reversely, and the method is favorable for accurately, rapidly and fully measuring the road surface evenness. According to the invention, the road flatness state RQI and the GPS are combined, the RQI and the GPS can be uploaded to a cloud server in real time during detection, the fine measurement of road surface diseases is realized, meanwhile, the detection result can be fused and analyzed with multi-source large data such as traffic volume, weather and park distribution, and the distribution rule of the road surface diseases is visually shown.
In addition, the invention can also automatically detect the pavement diseases based on machine vision and artificial intelligence. The invention relates to a pavement disease automatic detection based on machine vision and artificial intelligence, which is mainly completed by the cooperation of four sub-modules:
(1) high-speed industrial digital cameras: acquiring a road list photo in a self-adaptive manner;
(2) the vehicle-mounted terminal: the method comprises the steps of intelligently identifying the diseases by collecting pictures, calculating a road surface condition index (PCI) of a road section, matching the PCI with GPS data, and uploading the PCI to a cloud;
(3) a server on the cloud: receiving, analyzing and storing data;
(4) visualization terminal (data publishing, report generation).
The operation control process of the pavement disease video acquisition module comprises the following steps:
1. installation of equipment
The camera is installed in the middle of the top end of the vehicle, the lens faces the ground below the rear of the vehicle, no object is sheltered in a picture collected by the camera, and the picture is centered. The vehicle-mounted terminal, the transmission module and the positioning module are also installed in place.
2. Camera calibration
In order to ensure the accuracy of image acquisition and calculation, the visual angle and distortion of the image acquired by the camera need to be corrected by using a checkerboard method regularly according to the internal parameters of the camera lens. The calibration procedure used is as follows:
(1) a checkerboard calibration cloth is placed at a proper position in the middle of the picture and the picture is shot, and fig. 2 is a checkerboard picture collected in an initial test.
(2) Then find the corner points of the checkerboard and find the four corner points at the extreme edge (the four corner points forming the outermost rectangle of the checkerboard).
(3) The vertical lines of the rectangular area surrounded by these four corner points (vertical lines are defined here as lines ranging from-45 degrees to 45 degrees from the true north direction) are automatically found and considered only, as shown in fig. 3.
(4) According to the projection principle of camera shooting, all vertical straight lines are finally converged at one point to find the point.
(5) The convergence point is respectively connected with two angular points of the upper edge of the image, the extension line of each connecting line is intersected with the extension line of the lower edge of the image at one point, the two angular points of the upper edge of the image and the two intersection points of the extension line of the lower edge form a trapezoid, the trapezoid comprises the whole image, and the trapezoid represents the projection visual angle of the whole image.
(6) The projected visual angle is converted into a overlooked visual angle, namely two inclined edges of a trapezoid are converted into two parallel lines intuitively. However, in order to ensure that the ratio of the horizontal side and the vertical side of the checkerboard still maintains the actual ratio, the distance between the upper parallel side and the lower parallel side of the trapezoid also needs to be changed. That is, in the image with the viewing angle changed, the sizes of both the horizontal and vertical sides are changed from those of the original image.
(7) Finding the angular points of the checkerboard after the angle is converted, calculating the horizontal pixel distance between the two angular points, and obtaining the pixel number corresponding to the horizontal distance of one checkerboard in the top view by using the number of the checkerboard between the two angular points, and similarly, obtaining the pixel number corresponding to the vertical distance of one checkerboard in the top view. And corresponding to the actual size of the checkerboard, the actual size corresponding to a horizontal pixel distance and the actual size corresponding to a vertical pixel distance in the top view can be known.
(8) From the correspondence between the pixels obtained in step (7) and the actual dimensions of the calibration checkerboard, the actual dimensions or areas corresponding to the road surface damage in the picture taken at this viewing angle can be known, and the road surface areas corresponding to the whole picture taken can also be obtained, as shown in fig. 4.
3. Software startup
After the camera is calibrated, the vehicle-mounted terminal is started, and the image acquisition software and the transmission software in the vehicle-mounted terminal are automatically started.
4. Image acquisition
(1) The positioning acquisition sensor continuously acquires the position information of the vehicle and continuously calculates the assumed accumulated displacement of the vehicle; starting the vehicle from rest, and accumulating the displacement from 0;
(2) the industrial camera collects road surface pictures, and the frequency of collecting the road surface pictures can be adjusted according to the vehicle speed. When the vehicle speed is 0, the image acquisition frequency is 0; when the vehicle speed is 3.6km/h, the picture acquisition frequency is 1 Hz; when the vehicle speed is v km/h, the collection frequency of the pictures is (20 x v)/3.6 Hz.
(3) In the detection process, the acquired image can be stored in a local folder, the position information acquired by the positioning sensor can be stored in a local GPS database, and the acquired image path corresponding to the GPS can be written into the GPS database.
(4) When the accumulated displacement reaches or exceeds a preset distance, the vehicle-mounted terminal inputs all the images collected in the road section into a neural network algorithm, identifies the diseases of the ith type of road surface through the neural network algorithm and calculates the area A of the disease areaiCorresponding degree of disease wiAnd according to the 'road technical condition evaluation standard', calculating the road surface damage rate DR of the road sectioniAnd calculating the road surface condition index PCI of the road sectioniAnd meanwhile, the centroid position of the road section is calculated according to the position information acquired by the positioning sensor. The calculation result of the road surface condition index PCI and the corresponding road section centroid position are stored in a local road surface condition index database.
5. Pavement disease identification
And a disease recognition program in the vehicle-mounted terminal can continuously access a local image database, and when pictures which are collected by a camera and are not subjected to disease recognition processing exist in the image database, the pictures are calculated and processed according to a time sequence. The processing procedure is specifically shown in fig. 5, and includes the following steps:
(1) and (3) carrying out visual angle conversion processing on the photo according to the conversion relation of the photo visual angle and distortion to the photo in the step (2), wherein the effective area after the photo visual angle is changed is A.
(2) The disease type in the effective area of the picture after the visual angle transformation is identified by using the trained neural network, and the type A of the ith type of pavement disease in the picture can be identifiediCorresponding disease degree wiAnd calculating the road surface damage rate according to the following formula.
Wherein DR is the pavement damage rate, and is the percentage (%) of the sum of the damaged areas of various damages and the effective area in the photograph; w is aiAnd taking a value according to the evaluation standard of the technical condition of the highway as the weight of the i-th road surface damage.
(3) The vehicle-mounted terminal calculates the PCI of the road section according to the accumulated displacement, and specifically comprises the following steps: assuming that the vehicle is started from a static state, the displacement is accumulated from 0, when the accumulated displacement reaches or exceeds 20m, the road surface damage condition index (PCI) in the accumulated journey is calculated, the calculation result and the corresponding GPS data are recorded in a local PCI database, and the accumulated displacement of the vehicle-mounted terminal is accumulated from 0 again. The PCI calculation formula is as follows:
PCI＝100-a0DRa1
wherein DR is a road surface breakage rate, a0、a1The road surface type coefficients are respectively taken according to the evaluation standards of the technical conditions of the roads.
6. Data upload
The data in the PCI database are written into a local SD card or a hard disk of the vehicle-mounted terminal, meanwhile, in the detection process, the vehicle-mounted terminal system can continuously access a local GPS database, a flatness database and a road surface condition index database, and when records which are not uploaded exist in the database, the vehicle-mounted terminal system can call the data out and upload the data to the cloud-end platform through a transmission module (3G/4G/5G).
7. Data summarization and visual display
The cloud platform matches the pavement condition index PCI with the pavement evenness data and the GPS data, converts the international evenness index IRI into a pavement driving quality index RQI according to the 'assessment standard of road technical conditions' and matches the pavement driving quality index RQI with the GPS data, and meanwhile, the PCI and the RQI further match electronic road sections in a GIS map, so that the pavement evenness and the pavement damage condition can be visually displayed. After receiving a connection request of a remote vehicle-mounted terminal, the on-cloud server verifies the terminal identity, and if the terminal is a legal terminal, a data receiving thread is established for the terminal to receive data; and if the terminal is an illegal terminal, closing the link. And importing the received data into a GIS layer, matching the data with a corresponding road section, and visually displaying the PCI distribution condition of the road network through a system platform.
The method creatively uses the image visual angle and the distortion correction in the pavement disease image identification, and is favorable for accurately and quantitatively counting the pavement diseases. According to the invention, the PCI and the GPS are combined, and the detection result can be uploaded to the cloud server in real time, so that the fine measurement of the pavement diseases is realized, the relation among data such as the disease types, the occurrence positions and the disease quantity is established, and the distribution rule of the pavement diseases is visually shown. The invention applies the neural network technology and relies on lightweight equipment, and can realize the rapid and accurate identification of pavement diseases.
The invention has the following beneficial effects:
1. the system provided by the invention is based on lightweight equipment, full coverage, high-frequency and quick inspection of roads at all levels can be realized, the measured road surface running quality index RQI is combined with a GIS map, and the development rule of the road surface evenness can be displayed from the angles of time, space and the like; the detected road pavement damage condition PCI is combined with a GIS map, and the development rules of road diseases and road health conditions can be displayed from the angles of time, space and the like.
2. After the measured road surface running quality index RQI and the measured road surface condition index PCI are led into a GIS map, the visual display of the road surface evenness and the road surface damage condition can be provided for a road management and maintenance unit so as to be convenient for reference when a maintenance plan is made; the method can be fused and analyzed with multiple big data such as rainfall, air temperature, traffic volume, park distribution and the like, and is beneficial to estimating road pavement conditions, further to analyzing road disease laws, effectively promoting data-driven road maintenance decisions, improving the overall quality of road networks, and providing reliable theories and data support for road maintenance decisions, reasonable maintenance fund distribution and the like.
3. The road network level road surface flatness data can be provided for a road surface health platform or enterprise map navigation software, and provides each road surface flatness condition for travelers when planning a path, thereby optimizing path selection and improving the comfort level of drivers and passengers in the driving process.
Although the present invention has been described with respect to the preferred embodiments, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined by the appended claims.
Claims (7)
1. A road surface health condition rapid detection system is characterized by comprising:
an acquisition module: the method is used for acquiring vibration acceleration, pavement damage videos and GPS positioning data in the vehicle;
the vehicle-mounted terminal: the cloud server is used for receiving the data of each acquisition module, preprocessing the acquired data and uploading the processed data to the cloud server;
a server on the cloud: the system is used for receiving, analyzing and storing data uploaded by the vehicle-mounted terminal; and
visual terminal: the system is used for publishing data and generating reports, and providing real-time, multi-dimensional and customized display detection results;
the acquisition module is connected with the vehicle-mounted terminal through the transmission module, and the vehicle-mounted terminal, the cloud server and the visual terminal are connected through a network;
the acquisition module comprises an in-vehicle vibration acceleration acquisition module, a pavement damage video acquisition module and a GPS positioning acquisition module;
the pavement disease video acquisition module is an industrial digital camera, the resolution of the industrial digital camera is more than or equal to 200 ten thousand pixels, the maximum pixel is less than or equal to 7.5 mu m, and the aperture is more than or equal to F1.4;
the pavement disease video acquisition module is used for acquiring pavement pictures, identifying pavement diseases in the pictures through a deep neural network algorithm, and calculating a pavement condition index PCI according to the pavement diseases, and specifically comprises the following steps:
continuously acquiring the position information of the vehicle through a positioning acquisition sensor, and continuously calculating the assumed accumulated displacement of the vehicle; starting the vehicle from rest, and accumulating the displacement from 0;
the industrial digital camera collects road surface pictures, and the frequency of collecting the road surface pictures can be adjusted according to the vehicle speed; when the vehicle speed is 0km/h, the frequency of image acquisition is 0 Hz; when the vehicle speed is 3.6km/h, the picture acquisition frequency is 1 Hz; when the vehicle speed is v km/h, the acquisition frequency of the pictures is
In the detection process, the acquired image is stored in a local folder, the position information acquired by the positioning sensor is stored in a local GPS database, and the acquired image path corresponding to the GPS is written into the GPS database;
when the accumulated displacement reaches or exceeds a preset distance, the vehicle-mounted terminal inputs all the images collected in the road section into a neural network algorithm, identifies the diseases of the ith type of road surface through the neural network algorithm and calculates the area A of the disease areaiCorresponding degree of disease wiCalculating the road surface damage rate DR of the road sectioniAnd calculating the road surface condition index PCI of the road sectioniMeanwhile, the centroid position of the road section is calculated according to the position information acquired by the positioning sensor; the calculation result of the road surface condition index PCI and the corresponding road section centroid position are stored in a local road surface condition index database.
2. The system for rapidly detecting the road surface health condition according to claim 1, wherein the in-vehicle vibration acceleration acquisition module is two symmetrically distributed three-axis MEMS vibration sensors, and the two three-axis MEMS vibration sensors are fixed on a trunk of the vehicle and located above the rear wheels, so that an x-axis of the vibration sensor is parallel to a vehicle body direction, and a z-axis of the vibration sensor is parallel to a vehicle height direction.
3. A pavement health condition rapid detection method, which adopts the pavement health condition rapid detection system as claimed in claim 1, characterized in that the in-vehicle vibration acceleration acquisition module is two symmetrically distributed three-axis MEMS vibration sensors, the pavement disease video acquisition module is an industrial digital camera, and the vehicle-mounted terminal stores the acquired and processed data into a local GPS database, a flatness database and a pavement condition index database; the detection method comprises the following steps:
s1) road surface evenness collection: the method comprises the following steps that two MEMS vibration sensors acquire vibration information of a vehicle, and a vehicle-mounted terminal calculates an international road surface flatness index IRI by using a power spectral density algorithm;
s2) road surface disease condition collection: collecting a pavement picture by a vehicle-mounted industrial digital camera, identifying pavement diseases in the picture through a depth neural network algorithm, and calculating a pavement condition index PCI according to the pavement diseases;
the step S2) includes:
s21) the positioning acquisition sensor continuously acquires the position information of the vehicle and continuously calculates the assumed accumulated displacement of the vehicle; starting the vehicle from rest, and accumulating the displacement from 0;
s22) the industrial digital camera collects road surface pictures, and the frequency of collecting the road surface pictures can be adjusted according to the vehicle speed; when the vehicle speed is 0km/h, the frequency of image acquisition is 0 Hz; when the vehicle speed is 3.6km/h, the picture acquisition frequency is 1 Hz; when the vehicle speed is v km/h, the acquisition frequency of the pictures is
S23), in the detection process, the collected image is stored in a local folder, the position information collected by the positioning sensor is stored in a local GPS database, and the collected image path corresponding to the GPS is written into the GPS database;
s24) when the accumulated displacement reaches or exceeds a preset distance, the vehicle-mounted terminal inputs all the images collected in the road section into a neural network algorithm, identifies the disease of the ith type of road surface through the neural network algorithm and calculates the area A of the disease areaiCorresponding degree of disease wiCalculating the road surface damage rate DR of the road sectioniAnd calculating the road surface condition index PCI of the road sectioniMeanwhile, the centroid position of the road section is calculated according to the position information acquired by the positioning sensor; the PCI calculation result of the road condition index and the corresponding road section centroid position are stored in a local road condition index database;
s3) data uploading: in the detection process, the vehicle-mounted terminal continuously accesses a local GPS database, a flatness database and a road surface condition index database, and when records which are not uploaded exist in the databases, the vehicle-mounted terminal calls out the data and uploads the data to the cloud-end platform through the transmission module;
s4) visually displaying: the cloud server matches the road surface condition index PCI, the road surface smoothness data and the GPS data, converts the international smoothness index IRI into a road surface running quality index RQI and matches the road surface running quality index RQI with the GPS data, and further matches the road surface condition index PCI and the road surface running quality index RQI with an electronic road section in a GIS map, so that the road surface smoothness and the road surface damage condition can be visually displayed.
4. The method for rapidly detecting the road surface health condition according to claim 3, wherein the step S1) includes:
s11) the GPS positioning acquisition module continuously acquires the position information of the vehicle and continuously calculates the assumed accumulated displacement of the vehicle; starting the vehicle from rest, and accumulating the displacement from 0;
s12) continuously acquiring vibration data of the vehicle by the two vehicle-mounted MEMS vibration sensors, and transmitting the data to the vehicle-mounted terminal; the flatness data is collected at a frequency based onVehicle speed adjustment: when the vehicle speed is 0, the acquisition frequency is 0 Hz; when the vehicle speed is 3.6km/h, the acquisition frequency is 20 Hz; when the vehicle speed is 36km/h, the acquisition frequency is 200 Hz; when the vehicle speed is vkm/h, the acquisition frequency is
S13) when the accumulated displacement reaches or exceeds the preset distance, the vehicle-mounted terminal stores the vibration data collected all the time into a section of data to be processed, processes the data by using a power spectral density algorithm, and calculates the international flatness index IRI of the sectioniMeanwhile, the centroid position of the road section is calculated according to the position information collected by the positioning sensor, and the flatness calculation result and the corresponding road section centroid position are stored in a local flatness database;
s14), the accumulated displacement is reset to zero, and the steps S12 and S13 are repeated to start the position acquisition of the next link.
5. The method for rapidly detecting the road health condition according to claim 4, wherein the step S1) further comprises periodically performing a calibration test for the detected vehicle, and adjusting the relevant parameters of the power spectrum algorithm, wherein the calibration test is performed by: international flatness index IRI for a vehicle at a certain length0When the vehicle runs on a known road section, the vibration sensor collects the vibration information of the vehicle and calculates IRI through a power spectral density algorithm1And with IRI0And (3) comparison: if the result satisfies
The vehicle vibration parameter calibration result is considered to be accurate; otherwise, the vibration parameters of the vehicle are continuously adjusted until the vibration parameters are met
When calibration collection is carried out, the vehicle speeds are respectively controlled to be 20km/h, 40km/h and 60km/h, the vehicle runs at a constant speed, and at least 2 times of collection are carried out under each vehicle speed condition.
6. The method for rapidly detecting the road health condition according to claim 3, wherein the road pictures in the step S3) are collected as follows:
s31) carrying out visual angle conversion processing on the picture according to the conversion relation of the picture visual angle and distortion to the picture in the camera internal reference calibration process, wherein the effective area after the picture visual angle is changed is A;
s32) identifying the disease type in the effective area of the picture after the visual angle transformation by using the trained neural network, and identifying the type A of the ith type of pavement disease in the pictureiCorresponding disease degree wiAnd calculating the road surface damage rate according to the following formula:
in the formula, DR is the pavement damage rate and is the percentage of the sum of the whole damage area caused by various damages and the effective area in the picture; w is aiWeighting the damage of the i-th road surface;
s33) the vehicle-mounted terminal calculates the PCI of the road segment according to the accumulated displacement, specifically: assuming that the vehicle starts from a standstill, the displacement is accumulated from 0, when the accumulated displacement reaches or exceeds 20m, the road surface damage condition index PCI in the accumulated journey is calculated, the calculation result and the corresponding GPS data are recorded in a local PCI database, and meanwhile, the accumulated displacement of the vehicle-mounted terminal is accumulated again from 0; the PCI calculation formula is as follows:
PCI＝100-a0DRa1
wherein DR is a road surface breakage rate, a0、a1Respectively road surface type coefficients;
s34) the accumulated displacement is reset to zero, and the step S33) is repeated, and calculation of the road surface condition index of the next road section is started.
7. The method for rapidly detecting the road surface health condition according to claim 6, wherein the step S3) further comprises the following correction of the acquired image viewing angle and distortion:
s301) placing checkerboard calibration cloth at a proper position in the middle of a picture and shooting the picture;
s302) searching the angular points of the checkerboard, and searching four angular points at the outermost edge;
s303) automatically finding and only considering the vertical straight line of the rectangular area surrounded by the four corner points;
s304) according to the projection principle of camera shooting, each vertical straight line is finally converged at one point to find the point;
s305) respectively connecting a convergence point with two angular points of the upper edge of the image, wherein the extension line of each connecting line is intersected with the extension line of the lower edge of the image at a point, and the two angular points of the upper edge of the image and the two intersection points on the extension line of the lower edge form a trapezoid which comprises the whole picture and represents the projection angle of the whole picture;
s306) converting the projection visual angle into a overlooking visual angle, converting two inclined edges of the trapezoid into two parallel lines, and controlling the distance between an upper parallel edge and a lower parallel edge of the trapezoid to enable the proportion of the horizontal edge and the vertical edge of the checkerboard to still keep the actual proportion;
s307) searching angular points of the checkerboard after the angle of view is converted, calculating the transverse pixel distance between the two angular points and the number of the checkerboard between the two angular points, thereby obtaining the pixel number corresponding to the transverse distance/vertical distance of one checkerboard in a top view; corresponding to the actual size of the checkerboard, and acquiring the actual size corresponding to the horizontal pixel distance/vertical pixel distance in the top view;
s308) acquiring the actual size or area corresponding to the road surface damage in the picture shot at the visual angle according to the corresponding relation between the pixels acquired in the step S307) and the actual size of the calibration checkerboard, and further acquiring the road surface area corresponding to the whole shot picture.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910105679.3A CN109870456B (en) | 2019-02-01 | 2019-02-01 | Rapid detection system and method for road surface health condition |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910105679.3A CN109870456B (en) | 2019-02-01 | 2019-02-01 | Rapid detection system and method for road surface health condition |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109870456A CN109870456A (en) | 2019-06-11 |
| CN109870456B true CN109870456B (en) | 2022-01-28 |
Family
ID=66918601
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910105679.3A Active CN109870456B (en) | 2019-02-01 | 2019-02-01 | Rapid detection system and method for road surface health condition |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109870456B (en) |
Families Citing this family (40)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110930731A (en) * | 2019-10-08 | 2020-03-27 | 南京理工大学 | Intelligent transportation card and road surface detection system based on same |
| CN111638229B (en) * | 2019-10-08 | 2023-07-18 | 上海大学 | Airport pavement disease and satellite map bidirectional positioning correlation method based on Web |
| CN110969103B (en) * | 2019-11-21 | 2021-05-11 | 同济大学 | Method for measuring length of highway pavement disease based on PTZ camera |
| CN110942010B (en) * | 2019-11-22 | 2023-06-02 | 斑马网络技术有限公司 | Detection method and device for determining road surface fluctuation state based on vehicle-mounted camera |
| CN111160735A (en) * | 2019-12-16 | 2020-05-15 | 宁波市市政设施中心 | Bridgehead vehicle-jumping detection method based on LSTM recurrent neural network |
| CN111507233B (en) * | 2020-04-13 | 2022-12-13 | 吉林大学 | Multi-mode information fusion intelligent vehicle pavement type identification method |
| CN111610191B (en) * | 2020-04-20 | 2021-03-23 | 武汉理工大学 | Road detection and repair system |
| CN111785060A (en) * | 2020-06-18 | 2020-10-16 | 杭州恒领科技有限公司 | Special vehicle driving track information acquisition system based on 5G communication |
| CN111985494B (en) * | 2020-07-07 | 2023-12-22 | 深圳大学 | Road disease detection and classification method based on lightweight vehicle-mounted terminal |
| CN112067629A (en) * | 2020-09-09 | 2020-12-11 | 镇江千里路智能交通科技有限公司 | Vehicle-mounted lightweight pavement performance detection device |
| CN112505142B (en) * | 2020-10-27 | 2023-01-24 | 北京建筑大学 | Method for detecting damage of road structure, autonomous mobile device and storage medium |
| CN114518094A (en) * | 2020-11-16 | 2022-05-20 | 阿里巴巴集团控股有限公司 | Road detection method and system |
| CN112726351B (en) * | 2020-12-30 | 2021-08-27 | 上海同陆云交通科技有限公司 | Vehicle-mounted portable lightweight intelligent inspection method and system |
| CN112785867A (en) * | 2020-12-30 | 2021-05-11 | 上海智能交通有限公司 | Intelligent vehicle-mounted system for road health inspection |
| CN112800911B (en) * | 2021-01-20 | 2022-12-16 | 同济大学 | Pavement damage rapid detection and natural data set construction method |
| CN113012427B (en) * | 2021-02-09 | 2022-05-24 | 上海同陆云交通科技有限公司 | Improved vehicle-mounted lightweight inspection system and method |
| CN112985311B (en) * | 2021-02-09 | 2022-05-24 | 上海同陆云交通科技有限公司 | Vehicle-mounted portable lightweight intelligent inspection method and system |
| CN112907599B (en) * | 2021-02-22 | 2022-02-15 | 佳木斯大学 | Measuring device for indoor design |
| CN113155079B (en) * | 2021-03-01 | 2023-07-18 | 北京市市政工程研究院 | Road surface driving comfort judging method and device |
| EP4302081A1 (en) * | 2021-03-05 | 2024-01-10 | Telefonaktiebolaget LM Ericsson (publ) | Method and apparatus for road inspection |
| CA3211954A1 (en) * | 2021-03-15 | 2022-09-22 | Emil Sylvester RAMOS | System and method for automatic monitoring of pavement condition |
| CN113191269B (en) * | 2021-04-30 | 2023-06-09 | 南京金智视讯技术有限公司 | Road surface road condition real-time monitoring method based on 5G+AI intelligent analysis |
| CN113191291A (en) * | 2021-05-10 | 2021-07-30 | 南京金智视讯技术有限公司 | Road flatness detection method based on motion camera |
| CN113295137A (en) * | 2021-05-21 | 2021-08-24 | 吉林大学 | Road unevenness identification method based on vehicle wheel speed fluctuation |
| CN113592784A (en) * | 2021-07-08 | 2021-11-02 | 浙江科技学院 | Method and device for detecting pavement diseases based on lightweight convolutional neural network |
| CN113588664A (en) * | 2021-08-02 | 2021-11-02 | 安徽省通途信息技术有限公司 | Vehicle-mounted road defect rapid inspection and analysis system |
| CN113638294A (en) * | 2021-08-17 | 2021-11-12 | 广东粤高智慧交通科技有限公司 | Intelligent detection system technology for road maintenance diseases |
| CN113552138A (en) * | 2021-08-25 | 2021-10-26 | 浙江盘盘科技有限公司 | Cable drum quality detection system based on computer vision |
| CN113847944B (en) * | 2021-09-14 | 2024-02-02 | 石家庄铁道大学 | Road disease detection device and method based on reinforced learning memory control |
| CN113932758B (en) * | 2021-09-15 | 2022-12-20 | 同济大学 | Road surface flatness prediction method and device |
| CN113962301B (en) * | 2021-10-20 | 2022-06-17 | 北京理工大学 | Multi-source input signal fused pavement quality detection method and system |
| CN114996373B (en) * | 2022-03-31 | 2023-08-08 | 重庆市凤筑科技有限公司 | Public transportation big data system and method based on geographic information system and storage medium |
| CN114581856B (en) * | 2022-05-05 | 2022-08-09 | 广东邦盛北斗科技股份公司 | Agricultural unit motion state identification method and system based on Beidou system and cloud platform |
| CN115201218A (en) * | 2022-07-13 | 2022-10-18 | 鲁朗软件(北京)有限公司 | Vehicle-mounted pavement disease intelligent detection method and system |
| CN115713855A (en) * | 2022-10-14 | 2023-02-24 | 港珠澳大桥管理局 | Vehicle-mounted terminal system, data display method and device and computer equipment |
| CN115798194B (en) * | 2022-11-01 | 2024-01-19 | 西部科学城智能网联汽车创新中心(重庆)有限公司 | Road condition monitoring method and device based on vehicle vibration data |
| CN116087198B (en) * | 2022-12-02 | 2024-02-23 | 河南交通发展研究院有限公司 | Highway road surface situation data acquisition equipment and automatic rapid detection system thereof |
| CN116448016A (en) * | 2023-04-26 | 2023-07-18 | 成都智达万应科技有限公司 | Intelligent rapid detection system and detection vehicle with same |
| CN116448773B (en) * | 2023-06-19 | 2023-08-18 | 河北工业大学 | Pavement disease detection method and system with image-vibration characteristics fused |
| CN116972906B (en) * | 2023-06-21 | 2024-02-13 | 清华大学 | Highway state evaluation method, device, storage medium and collection box |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104790283A (en) * | 2015-04-10 | 2015-07-22 | 同济大学 | Quick road surface roughness detection system based on vehicle-mounted accelerometer |
| CN104864878A (en) * | 2015-05-22 | 2015-08-26 | 汪军 | Electronic map based road condition physical information drawing and inquiring method |
| CN106529593A (en) * | 2016-11-08 | 2017-03-22 | 广东诚泰交通科技发展有限公司 | Pavement disease detection method and system |
| DE102017001364A1 (en) * | 2017-02-14 | 2017-10-19 | Daimler Ag | Method for environment detection and visualization |
| CN108335002A (en) * | 2017-01-20 | 2018-07-27 | 亚信蓝涛（江苏）数据科技有限公司 | A kind of visual road maintenance big data analysis system |
-
2019
- 2019-02-01 CN CN201910105679.3A patent/CN109870456B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104790283A (en) * | 2015-04-10 | 2015-07-22 | 同济大学 | Quick road surface roughness detection system based on vehicle-mounted accelerometer |
| CN104864878A (en) * | 2015-05-22 | 2015-08-26 | 汪军 | Electronic map based road condition physical information drawing and inquiring method |
| CN106529593A (en) * | 2016-11-08 | 2017-03-22 | 广东诚泰交通科技发展有限公司 | Pavement disease detection method and system |
| CN108335002A (en) * | 2017-01-20 | 2018-07-27 | 亚信蓝涛（江苏）数据科技有限公司 | A kind of visual road maintenance big data analysis system |
| DE102017001364A1 (en) * | 2017-02-14 | 2017-10-19 | Daimler Ag | Method for environment detection and visualization |
Non-Patent Citations (2)
| Title |
|---|
| 公路路面裂缝检测与识别技术研究;王婷婷;《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》;20180115(第01期);第4-5，7，45-59页 * |
| 鱼眼畸变棋盘格图像校正;黄岩岩等;《计算机工程与应用》;20141231;第50卷(第12期);第111-114页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109870456A (en) | 2019-06-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109870456B (en) | Rapid detection system and method for road surface health condition | |
| CN109716108B (en) | Bituminous paving disease detecting system based on two mesh image analysis | |
| CN105905132B (en) | A kind of track condition intelligent patrol detection apparatus and method | |
| CN104567708B (en) | Full section of tunnel high speed dynamical health detection means and method based on active panoramic vision | |
| CN205632521U (en) | Track state intelligence inspection device | |
| CN106049210B (en) | A kind of track condition Intelligent Measurement platform | |
| WO2022206978A1 (en) | Roadside millimeter-wave radar calibration method based on vehicle-mounted positioning apparatus | |
| CN101694084B (en) | Ground on-vehicle mobile detecting system | |
| CN103778681A (en) | Vehicle-mounted high-speed road inspection system and data acquisition and processing method | |
| CN112836737A (en) | Roadside combined sensing equipment online calibration method based on vehicle-road data fusion | |
| CN106978774B (en) | A kind of road surface pit slot automatic testing method | |
| CN106296814B (en) | Highway maintenance detection and virtual interactive interface method and system | |
| CN105654732A (en) | Road monitoring system and method based on depth image | |
| JP2019196680A (en) | Pavement information gathering inspection system, pavement information gathering inspection method, and program | |
| CN108107462A (en) | The traffic sign bar gesture monitoring device and method that RTK is combined with high speed camera | |
| CN112985311B (en) | Vehicle-mounted portable lightweight intelligent inspection method and system | |
| CN103837087A (en) | Automatic pantograph detection method based on active shape model | |
| CN113588664A (en) | Vehicle-mounted road defect rapid inspection and analysis system | |
| CN115506216A (en) | Pavement evenness analysis method and maintenance inspection system | |
| CN114971166A (en) | Lightweight road pavement service quality inspection system | |
| TW200925354A (en) | Robot and method for automatically detecting flatness and surface breakage | |
| CN116448126A (en) | Automatic compensation and verification system for navigation electronic map data generation | |
| US11483485B2 (en) | Device for collecting photos of field surface feature and information measurement and calculation method | |
| CN111858975B (en) | Online query method for railway vehicle-mounted laser scanning panoramic image | |
| JP2019039850A (en) | Railway facility monitoring device and railway facility monitoring system |
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 | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20211122 Address after: 200082 4th floor, block a, 1229 Zhoujiazui Road, Yangpu District, Shanghai Applicant after: SHANGHAI INTELLIGENT TRANSPORTATION Co.,Ltd. Applicant after: Shanghai urban construction digital industry group Co., Ltd Address before: 200000 room A107, first floor, building 64, No. 1436, Jungong Road, Yangpu District, Shanghai Applicant before: SHANGHAI INTELLIGENT TRANSPORTATION Co.,Ltd. |
|
| TA01 | Transfer of patent application right | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |