Background
The rural highway plays an extremely important role in the process of promoting economic and social progress, is used as one of main traffic outgoing carriers for connecting rural areas with the outside, and can drive the development of rural economy. Along with the promotion of the strategy of village promotion in recent years, the construction scale of various rural basic services is continuously increased, and the effect is remarkable particularly in the aspect of road construction. However, in view of the current practical situation, various diseases often occur in the process of putting rural highways into operation in the later period, which greatly affects the use functions and service lives of the rural highways, and the damage to the highways is continuously increased along with the trend that vehicles in rural areas rise year by year, so that the analysis of road surface diseases is necessary, and matched maintenance measures are taken, thereby protecting the safe operation of the rural highways.
The construction condition of 'four good rural villages' in the 'thirteen-five' period is that the country accumulatively completes the new reconstruction of 138.8 kilometers of rural highways, the total mileage of rural highways in the country reaches 420 kilometers, the total mileage of rural highways in Sichuan province reaches 34.7 kilometers, and the rural highways account for 83.8 percent of the total mileage of the highways. The rapid development of rural highways is reflected from the data, the highway maintenance cost is continuously increased, the investment of the project can meet the maintenance capital requirement of the rapid development of rural areas, and the maintenance development of the rural highways is greatly promoted. Therefore, we need to introduce new technologies and new means to enable rural highway management. Insist on the intelligent development, promote to apply new technology, new means enable rural highway management maintenance work. The application of new technologies such as 5G, Beidou, Internet of things, big data, satellite remote sensing and the like is strengthened, and the rural highway management efficiency and maintenance level are continuously improved.
The informatization level of rural highway management is continuously improved, and a 'quick detection technology' is also proposed for the first time; the method has the advantages of promoting the digitization of the basic information of the rural highway, strengthening the application of new technologies such as internet, satellite remote sensing, rapid detection and the like, establishing a long-acting mechanism for data acquisition and processing, perfecting the comprehensive supervision capability of the rural highway, strengthening the information resource sharing and exchange and improving the management efficiency of the rural highway.
For an auxiliary driving system of an intelligent automobile, the key point is that the method can accurately identify and detect a path target and a vehicle environment, and makes reasonable judgment through data analysis. Road conditions and driving environments are complex and variable, and generally include road signs, lane lines, traffic signs, obstacles, pedestrians, vehicles, traffic signals and the like. How to quickly and accurately detect a target path and a driving state in an effective time becomes a difficult point for researching the automobile safety auxiliary driving technology. Along with the rapid development of the vision technology, the machine vision is more accurate in target identification and detection, the quantity of acquired information is large, the method does not depend on the change of the road shape, multi-target detection can be performed, the maintenance cost is low, and the storage is convenient. Therefore, vision technology represents a great advantage in road identification and tracking of smart cars. The research of machine vision in the field of intelligent transportation has been carried out for a long time abroad, and related mature products come out, but the research in the aspect of China is relatively late, the technology is not mature enough, and the related products are not many. Therefore, research on intelligent car-assistant driving technology based on vision technology is urgently carried out.
The machine vision is to acquire images through a camera, send the images to a computer for processing, and acquire characteristic information contained in the images so as to realize target recognition and detection. With the rapid development of image processing technology, machine vision has been widely applied in various fields such as detection and monitoring, industrial production, visual navigation, human-computer interaction and the like, and great convenience is brought to the production and life of people. Especially in the field of intelligent automobiles, the application of vision technology brings great changes to the automobile revolution.
The defects and shortcomings of the prior art are as follows: 1. the GPS error of the vehicle-mounted equipment is large, and specific diseases cannot be accurately displayed on a map. 2. Road diseases are not reported in time, and the reported positions of the diseases are too many, so that the data volume is too large, invalid data are redundant, the reading and analyzing efficiency is completely influenced, and the visualization of data visual display is influenced. 3. The labor cost required by highway disease management is high, and the efficiency is low.
Therefore, there is a need to provide a new intelligent reporting system for road diseases to solve one or more of the above technical problems.
Disclosure of Invention
In order to solve one or more technical problems, the invention provides a deviation rectifying and road disease intelligent reporting system based on a GPS, which comprises a reporting terminal and an intelligent analysis cloud.
Specifically, the reporting terminal is deployed on mobile and/or vehicle-mounted intelligent equipment and establishes communication connection with the intelligent analysis cloud through a communication network; the reporting terminal can collect/upload GPS positioning data and road image data in real time and receive/upload data through a data packet and an intelligent analysis cloud.
More specifically, the intelligent analysis cloud is provided with a picture identification and classification unit, a GPS (global positioning system) deviation correction algorithm unit, a GPS basic database, a GIS network database and a GIS map visualization analysis system.
In detail, the image identification and classification unit is used for receiving the road image data uploaded by the report terminal and carrying out image identification and classification on the road image data to obtain a road disease identification result; the GPS deviation correction algorithm unit is used for receiving the GPS positioning data uploaded by the reporting terminal, and carrying out GPS route positioning and positioning deviation correction on the GPS positioning data to obtain an accurate positioning coordinate; the GPS basic database is used for storing basic data required by GPS positioning; the GIS network database is used for storing GIS network data; the GIS map visualization analysis system is used for visually displaying the spatial information map of the reported information of the reporting terminal.
As a further solution, the GPS basic database and the GIS network database both use mysql database as storage database and Redis database as cache database; the Redis database is deployed with Redis Geo and performs data processing through Redis message queue data.
As a further solution, the GPS deviation correction algorithm unit obtains the offset constant coefficient of the positioning road section by the following steps
:
S1, receiving the GPS positioning data uploaded by the report terminal;
s2 setting the latitude and longitude of the current location point as upload point P (X)0,Y0);
S3 finding nearby positioning point A (X) through Redis Geo1,Y1) The positioning point A belongs to the route L;
s4 matches any value of two forward and backward points through the positioning point A as the positioning point B (X)2,Y2);
S5 finds the vector AB (X) between the anchor A and the anchor B
2-X
1,Y
2-Y
1) Wherein, in the step (A),
;
s6, making a perpendicular line from the uploading point P to the L line, and setting the intersection point as a perpendicular point Q (X, Y);
s7 finds the upload point P and the vector PQ (X-X)
0,Y-Y
0) Wherein, in the step (A),
;
s8 is represented by vector PQ perpendicular to vector AB, i.e.: the mode of the projection of the vector PQ on the vector AB is 0, the dot product of the projection is 0, and the coordinate of a Q point is calculated; wherein the content of the first and second substances,
;
s9, judging whether the point Q is on the straight line L line, namely calculating the formula:
whether the result is true or not; if yes, marking the point Q as a deviation correcting point Q corresponding to the uploading point P;
s10 calculating the distance S between the positioning point A and the deviation correcting point Q
1Wherein, in the step (A),
;
s11, acquiring longitude and latitude of each point;
s12 calculates the actual distance S between locating point A longitude and latitude lat1, long 1 and uploading point P longitude and latitude lat2, long 22The calculation formula is as follows:
(ii) a Wherein R is the radius of the earth, namely R =6730KM, S
2Is the distance between two GPS points;
s13 reaction of S1And S2An offset is introduced, and an offset M of the upload point P is calculated, where M = S1/S2;
S14 repeating S1-S13 to obtain average value of multiple offsets of L line as regular offset constant coefficient
The formula is as follows:
(ii) a Wherein n is the measurement times, i is the measurement number, and the obtained offset constant coefficient
。
As a further solution, the GPS deviation correction algorithm unit judges whether the GPS positioning data deviates from the GIS network data by the following steps:
a1 making a vertical line from the uploading point P to the vector AB to obtain a vertical point Q;
a2 judges whether the vertical point Q is on the segment AB: i.e. formula of calculation
Whether the result is true or not; if yes, the vertical point Q is in the AB line segment; if the vertical point Q is not in the AB line segment, the vertical point Q is not in the AB line segment;
a3 using offset constant coefficient if vertical point Q is not in AB line segment
Calculating a coordinate point of the vertical point Q in the AB line segment; if the vertical point Q is in the AB line segment, directly calculating the coordinate of the vertical point Q, and executing the next step;
a4, searching whether the coordinate corresponding to the vertical point Q exists in a GPS basic database, wherein a GPS repetition threshold is preset, and if the distance between the two points is greater than the GPS repetition threshold, determining the two different coordinates; if the distance between the two points is not greater than the GPS repetition threshold, the two points are regarded as two identical coordinates;
a5, if the coordinate corresponding to the vertical point Q already exists in the GPS basic database, only updating the reported information corresponding to the coordinate; and if the coordinate corresponding to the vertical point Q does not exist in the GPS basic database, newly building and storing the coordinate into the GPS basic database, and newly building the report information corresponding to the coordinate.
As a further solution, the picture recognition and classification unit is an AI-based picture recognition and classification unit, and trains the model through a disease picture classification training library to obtain the characteristic values of each disease classification, and the steps are as follows:
b1, collecting road disease pictures and carrying out classification treatment to obtain a disease picture classification training library;
b2, inputting the disease picture classification training library into a picture recognition and classification unit according to classification and extracting features;
b3, obtaining the characteristics of the road diseases and calculating the characteristic values to obtain the classification characteristic values of the diseases;
b4, adding the disease classification characteristic values corresponding to various road diseases into a disease classification characteristic value database.
As a further solution, the picture identification and classification unit obtains the classification result by:
c1, receiving and reporting road image data uploaded by the terminal;
c2 extracting road image data characteristics through an image identification and classification unit and calculating disease image characteristic values;
c3, writing the characteristic value of the disease picture into a Redis database for caching;
c4, writing the disease classification characteristic value database into a Redis database for caching;
c5, comparing and analyzing the characteristic value of the disease picture with the characteristic value database of the disease classification;
c6, if the disease picture characteristic value is matched with the disease classification characteristic value database, outputting a disease classification result; and if the matching is not established, the disease classification result is the unidentified disaster.
As a further solution, the GIS map visualization analysis system is a layered, dynamic and delivery map visualization system; the GIS map visualization analysis system comprises a GIS road network layer, a disease layer and a patrol layer; the GIS road network map layer comprises routes, road sections, bridges, tunnels, culverts, side slopes, intercommunication areas, traffic implementation, traffic volume and data files; the disease layer shows specific positions by a disease chart, wherein the specific positions comprise a disease audit state, a disease disposal state, a disease site, disease classification, a reporting unit, a disease picture, a disease area and disease volume information; the patrol map layer is used for displaying daily patrol real-time tracks of patrol personnel, and comprises patrol tracks, patrol times and patrol mileage.
Compared with the prior art, the system for correcting the deviation and intelligently reporting the road diseases based on the GPS has the following beneficial effects:
1. the intelligent reporting system provided by the invention can identify the road surface condition and can report, check and process the diseases; maintenance work plans are reasonably arranged, labor cost is reduced, and output efficiency is improved;
2. the method can accurately display the disease position on the map by combining a GPS (global positioning system) deviation correction algorithm unit with a visualization technology, and accurately display the disease on the Gaode map, wherein the uploaded longitude and latitude and basic road data find out the nearest coordinate of the GPS position through the Redis Geo, the Redis Geo can conveniently calculate the point of the nearest distance, and then the deviation correction is performed through the deviation correction algorithm, so that the matching line which is completely matched with the GIS network database is realized, and the precise display on the Gaode map is convenient;
3. the GIS map visualization analysis system used by the invention is a layered, dynamic and delivered map visualization system, and comprises a GIS road network layer, a disease layer and an inspection layer; the user can intuitively, quickly, accurately and multi-level know the road disease condition; the working efficiency is improved;
4. the method comprises the steps of processing and analyzing reported diseases through a picture recognition and classification technology, simultaneously performing GPS (global positioning system) deviation correction processing on reported GPS longitude and latitude, reporting to an intelligent analysis cloud, and visually displaying specific positions of the diseases through a GIS (geographic information system) map visual analysis system; the inspection cost and the detection cost are reduced to a great extent, the road inspection efficiency is improved, and intelligent management such as intellectualization, automation, visualization and the like is realized.
Detailed Description
The invention is further described with reference to the following figures and embodiments.
As shown in fig. 1 to 4, in order to solve one or more of the above technical problems, the present embodiment provides a system for deviation correction and intelligent reporting of a road fault based on a GPS, which includes a reporting terminal and an intelligent analysis cloud.
Specifically, the reporting terminal is deployed on mobile and/or vehicle-mounted intelligent equipment and establishes communication connection with the intelligent analysis cloud through a communication network; the reporting terminal can collect/upload GPS positioning data and road image data in real time and receive/upload data through a data packet and an intelligent analysis cloud.
More specifically, the intelligent analysis cloud is provided with a picture identification and classification unit, a GPS (global positioning system) deviation correction algorithm unit, a GPS basic database, a GIS network database and a GIS map visualization analysis system.
In detail, the image identification and classification unit is used for receiving the road image data uploaded by the report terminal and carrying out image identification and classification on the road image data to obtain a road disease identification result; the GPS deviation correction algorithm unit is used for receiving the GPS positioning data uploaded by the reporting terminal, and carrying out GPS route positioning and positioning deviation correction on the GPS positioning data to obtain an accurate positioning coordinate; the GPS basic database is used for storing basic data required by GPS positioning; the GIS network database is used for storing GIS network data; the GIS map visualization analysis system is used for visually displaying the spatial information map of the reported information of the reporting terminal.
It should be noted that: the bottom layer main core technology of the intelligent reporting system based on GPS deviation correction and road diseases provided by the implementation adopts a GPS deviation correction technology and a disease picture identification and classification technology; the GPS deviation rectifying technology based on Redis Geo aims to solve the problems of inaccuracy and overlarge offset of the position of a damaged GPS uploaded by equipment. The disease reporting is mainly realized through a mobile phone app and a vehicle-mounted device AI, namely a mobile and/or vehicle-mounted intelligent device. The mobile phone app reports in a voice and picture mode, a picture recognition and classification technology is adopted to process and analyze reported diseases, GPS deviation correction processing is carried out on reported GPS longitude and latitude, intelligent road disease reporting is carried out on a road surface by a vehicle carrying AI algorithm device, and specific positions of the diseases are visually displayed through a Gade map layer. The inspection cost and the detection cost are reduced to a great extent, the road inspection efficiency is improved, and intelligent management such as intellectualization, automation, visualization and the like is realized.
The intelligent inspection equipment is installed on operating vehicles (such as buses, cleaning vehicles, rural passenger vehicles and the like) and law enforcement vehicles, and vehicle-mounted equipment AI is loaded to realize rapid detection and presentation of a target value of a detected road surface. The user can realize the short-term test to the road of process in step when not disturbing the daily operation of vehicle, and main detection objective is data such as road surface crack, road surface hole groove, road surface wave hug, road surface running quality index RQI to show the testing result adoption visual mode.
As a further solution, the GPS basic database and the GIS network database both use mysql database as storage database and Redis database as cache database; the Redis database is deployed with Redis Geo and performs data processing through Redis message queue data.
It should be noted that: the distance between the disease reported GPS data and the point on the L line can be obtained by utilizing the GEO geographic positioning calculation of redis, the longitude and latitude (coordinates) of the point on the L line stored in the GIS network database are calculated by the GEO to obtain the distance. And the GPS basic database and the GIS road network basic database are stored by adopting mysql database. The cache database adopts a Redis database. And simultaneously, the Redis message queue is used for data processing and the Redis Geo is used for calculating the longitude and the latitude. The GIS road network basic database is a GIS network database of a road basic database of a scientific research institute of the department of transportation, leads out GIS network data of a certain area, and obtains the GIS network database through data cleaning. Reference data for a GPS deskew algorithm.
As a further solution, as shown in FIG. 2, the GPS deviation correction algorithm unit obtains the offset constant coefficient of the positioning road section by the following steps
:
S1, receiving the GPS positioning data uploaded by the report terminal;
s2 setting the latitude and longitude of the current location point as upload point P (X)0,Y0);
S3 finding nearby through Redis GeoLocation point A (X)1,Y1) The positioning point A belongs to the route L;
s4 matches any value of two forward and backward points through the positioning point A as the positioning point B (X)2,Y2);
S5 finds the vector AB (X) between the anchor A and the anchor B
2-X
1,Y
2-Y
1) Wherein,
;
s6, making a perpendicular line from the uploading point P to the L line, and setting the intersection point as a perpendicular point Q (X, Y);
s7 finds the upload point P and the vector PQ (X-X)
0,Y-Y
0) Wherein, in the step (A),
;
s8 is represented by vector PQ perpendicular to vector AB, i.e.: the mode of the projection of the vector PQ on the vector AB is 0, the dot product of the projection is 0, and the coordinate of a Q point is calculated; wherein the content of the first and second substances,
;
s9, judging whether the point Q is on the straight line L line, namely calculating the formula:
whether the result is true or not; if yes, marking the point Q as a deviation correcting point Q corresponding to the uploading point P;
s10 calculating the distance S between the positioning point A and the deviation correcting point Q
1Wherein, in the step (A),
;
s11, acquiring longitude and latitude of each point;
s12 calculates the actual distance S between locating point A longitude and latitude lat1, long 1 and uploading point P longitude and latitude lat2, long 22The calculation formula is as follows:
(ii) a Wherein R is the radius of the earth, namely R =6730KM, S
2Is the distance between two GPS points;
s13 reaction of S1And S2An offset is introduced, and an offset M of the upload point P is calculated, where M = S1/S2;
S14 repeating S1-S13 to obtain average value of multiple offsets of L line as regular offset constant coefficient
The formula is as follows:
(ii) a Wherein n is the measurement times, i is the measurement number, and the obtained offset constant coefficient
。
As a further solution, as shown in fig. 2, the GPS deviation rectifying algorithm unit determines whether the GPS positioning data deviates from the GIS network data by the following steps:
a1 making a vertical line from the uploading point P to the vector AB to obtain a vertical point Q;
a2 judges whether the vertical point Q is on the segment AB: i.e. formula of calculation
Whether the result is true or not; if yes, the vertical point Q is in the AB line segment; if the vertical point Q is not in the AB line segment, the vertical point Q is not in the AB line segment;
a3 using offset constant coefficient if vertical point Q is not in AB line segment
Calculating a coordinate point of the vertical point Q in the AB line segment; if the vertical point Q is in the AB line segment, directly calculating the coordinate of the vertical point Q, and executing the next step;
here, it should be noted that: multiplying the distance from the uploading point P to the positioning point A by an offset constant coefficient
The distance from the vertical point Q to the point A in the AB line segment can be obtained, and the linear equation can be solved by knowing the coordinates of the two points A, B; the coordinate of the vertical point Q can be obtained by the simultaneous connection of the linear equation and the distance from the vertical point Q to the point A.
A4, searching whether the coordinate corresponding to the vertical point Q exists in a GPS basic database, wherein a GPS repetition threshold is preset, and if the distance between the two points is greater than the GPS repetition threshold, determining the two different coordinates; if the distance between the two points is not greater than the GPS repetition threshold, the two points are regarded as two identical coordinates;
a5, if the coordinate corresponding to the vertical point Q already exists in the GPS basic database, only updating the reported information corresponding to the coordinate; and if the coordinate corresponding to the vertical point Q does not exist in the GPS basic database, newly building and storing the coordinate into the GPS basic database, and newly building the report information corresponding to the coordinate.
It should be noted that: in the embodiment, the GPS deviation correction algorithm unit can be combined with the visualization technology to accurately display the disease position on the map and accurately display the disease on the Gade map. The GPS deviation correction step comprises a GPS matching line, GPS vertical point calculation, GPS constant coefficient calculation and GPS duplication removal. Firstly, the equipment for reporting diseases reports longitude and latitude information. The latest coordinates of the GPS position are found out through the uploaded longitude and latitude and basic road data through the Redis Geo, the Redis Geo can conveniently calculate the point of the closest distance, the number of the returned records is set to be limited to be 10, the more the number of the closest points matched to the line is, the equipment point is marked on the line, and then the deviation is corrected through a deviation correction algorithm, so that the matching line matched with the GIS network database is completely matched, and the accurate display on a Gaode map is facilitated.
As a further solution, as shown in fig. 3, the picture recognition and classification unit is an AI-based picture recognition and classification unit, and trains the model through a disease picture classification training library to obtain characteristic values of disease classifications, which includes the following steps:
b1, collecting road disease pictures and carrying out classification treatment to obtain a disease picture classification training library;
b2, inputting the disease picture classification training library into a picture recognition and classification unit according to classification and extracting features;
b3, obtaining the characteristics of the road diseases and calculating the characteristic values to obtain the classification characteristic values of the diseases;
b4, adding the disease classification characteristic values corresponding to various road diseases into a disease classification characteristic value database.
As a further solution, the picture identification and classification unit obtains the classification result by:
c1, receiving and reporting road image data uploaded by the terminal;
c2 extracting road image data characteristics through an image identification and classification unit and calculating disease image characteristic values;
c3, writing the characteristic value of the disease picture into a Redis database for caching;
c4, writing the disease classification characteristic value database into a Redis database for caching;
c5, comparing and analyzing the characteristic value of the disease picture with the characteristic value database of the disease classification;
c6, if the disease picture characteristic value is matched with the disease classification characteristic value database, outputting a disease classification result; and if the matching is not established, the disease classification result is the unidentified disaster.
As a further solution, the GIS map visualization analysis system is a layered, dynamic and delivery map visualization system; the GIS map visualization analysis system comprises a GIS road network layer, a disease layer and a patrol layer; the GIS road network map layer comprises routes, road sections, bridges, tunnels, culverts, side slopes, intercommunication areas, traffic implementation, traffic volume and data files; the disease layer shows specific positions by a disease chart, wherein the specific positions comprise a disease audit state, a disease disposal state, a disease site, disease classification, a reporting unit, a disease picture, a disease area and disease volume information; the patrol map layer is used for displaying daily patrol real-time tracks of patrol personnel, and comprises patrol tracks, patrol times and patrol mileage.
As shown in fig. 4, in this embodiment, one-key reporting and inspection of a road disease is realized through a mobile phone APP, and users in villages, towns and counties use mobile terminals to report various road condition information such as road sealing, water accumulation, construction and the like. The road diseases are divided into district and county and country/village roads according to areas. The management area is planned with planned road diseases every year, and the diseases are planned as the diseases of a plan type according to the disease condition of the management area and the disease treatment budget cost. Diseases that are not planned and severe are classified as emergent. It can be divided into a plan class and an emergency class.
And (5) reporting the road disease. Firstly, if the disease is in a rural/village road, the patrolman begins to report the disease in the plan to a traffic management station, the traffic management station plans a disease treatment plan book, and a maintenance manager of a disease area begins to treat the disease through the approval of the station leader; if the disease is an unplanned disease, a traffic management station draws a disease treatment plan book, and the station leader approves the approval and needs to approve the approval, and then a maintenance manager of a disease area starts to treat the disease. Secondly, if the disease is a county road, the patrol personnel report plan diseases, a treatment scheme is drawn up by a department of the engineering stock, the maintenance section is long and is subjected to one-level examination, the first-level examination is passed, the department of the management department is established and is subjected to examination and passing, the examination and passing are ended, finally, the department of the engineering stock in the disease area is used for treating the disease process, and the department of the supervision department or the principal and the subordinate of a service center are required to carry out examination and recheck in the treatment of the county road. After the examination is passed, the disease treatment is completed.
In conclusion, the GPS-based deviation correction and intelligent road disease reporting system realizes multi-element automatic identification such as basic diseases, various intersection installation and application identification and judgment. The application of the new rural highway technology is promoted, the intelligent application of the GPS deviation correction technology, the picture recognition and classification technology, the GIS map visualization technology and other technologies is realized, the scientific and technological management is realized, the rural highway management and maintenance work is enabled, the rural joy is promoted, the rural economic development is driven, and the good social benefit is produced.
The above description is only an embodiment of the present invention, and not intended to limit the scope of the present invention, and all modifications of equivalent structures and equivalent processes, which are made by using the contents of the present specification and the accompanying drawings, or directly or indirectly applied to other related technical fields, are included in the scope of the present invention.