CN112270689A - Road boundary remote measurement and identification algorithm adopting artificial neural network technology - Google Patents

Road boundary remote measurement and identification algorithm adopting artificial neural network technology Download PDF

Info

Publication number
CN112270689A
CN112270689A CN202011116151.5A CN202011116151A CN112270689A CN 112270689 A CN112270689 A CN 112270689A CN 202011116151 A CN202011116151 A CN 202011116151A CN 112270689 A CN112270689 A CN 112270689A
Authority
CN
China
Prior art keywords
boundary
road
scanning
axis
points
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.)
Withdrawn
Application number
CN202011116151.5A
Other languages
Chinese (zh)
Inventor
周芦慧
肖良军
刘利民
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Huzhou University
Original Assignee
Huzhou University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Huzhou University filed Critical Huzhou University
Priority to CN202011116151.5A priority Critical patent/CN112270689A/en
Publication of CN112270689A publication Critical patent/CN112270689A/en
Withdrawn legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/181—Segmentation; Edge detection involving edge growing; involving edge linking
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/60—Analysis of geometric attributes
    • G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10032—Satellite or aerial image; Remote sensing
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20084—Artificial neural networks [ANN]

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Health & Medical Sciences (AREA)
  • Computing Systems (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • Biomedical Technology (AREA)
  • Molecular Biology (AREA)
  • Biophysics (AREA)
  • General Engineering & Computer Science (AREA)
  • Artificial Intelligence (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Geometry (AREA)
  • Image Analysis (AREA)

Abstract

The invention provides a road boundary telemetering identification algorithm adopting an artificial neural network technology, which specifically comprises the following steps: s1, obtaining distance information between a scanning probe and a ground scanning point; s2, defining a longitudinal track of the scanning probe as a y axis, defining transverse scanning to the ground as an x axis, and defining a height difference between ground scanning points as a z axis; s3, calculating adjacent dot matrixes of different x-ray points of the same point on the y axis through inference calculation, obtaining a corresponding x-ray point position closest to the boundary at the position where sudden change occurs according to the change trend of the difference of the adjacent points on the z axis, and finding a series of x points along with the change of the y value to determine the initial value of the road boundary; and S4, carrying out smooth analysis and correction on the initial value of the road boundary to obtain the effective road boundary based on x and y coordinates. The method can achieve the purpose of automatically providing the measured boundary of the road, and provide scientific basis for determining the flatness, the area and the calculation range of other parameters of the road surface.

Description

Road boundary remote measurement and identification algorithm adopting artificial neural network technology
[ technical field ] A method for producing a semiconductor device
The invention relates to the technical field of road telemetering data analysis, in particular to a road boundary telemetering identification algorithm adopting an artificial neural network technology, and more particularly, the road telemetering data is analyzed through an artificial intelligent neural network technology to achieve the purpose of automatically determining an effective boundary of a road.
[ background of the invention ]
The current method for determining the road boundary has the following defects:
1. there is a lack of solutions for non-manual determination of road boundaries.
2. A solution for determining road boundaries by telemetry data is lacking.
3. And an executable scheme for automatically determining the road boundary through telemetering data analysis is not shown in the prior art, and artificial intelligence technology is not adopted.
[ summary of the invention ]
The invention aims to solve the problems in the prior art and provides a road boundary telemetering and identifying algorithm adopting an artificial neural network technology, which has the advantages of high processing speed, good universality and strong effectiveness, can save manpower and intelligently and automatically determine the road boundary.
In order to achieve the above object, the present invention provides a road boundary telemetry recognition algorithm using artificial neural network technology, which is based on artificial intelligent neural network and automatically obtains road effective boundary by performing inference calculation on field telemetry data, and specifically comprises the following steps:
s1, acquiring field telemetering data: acquiring distance information between a scanning probe and a ground scanning point;
s2, setting a three-dimensional coordinate system: defining a longitudinal track of a scanning probe as a y-axis, defining transverse scanning to the ground as an x-axis and defining the height difference between ground scanning points as a z-axis;
s3, determining an initial value of a road boundary: calculating corresponding x point positions closest to the boundary at positions where sudden changes occur according to the change trend of the difference of adjacent points on the z axis through the adjacent dot matrix reasoning calculation of different x line points of the same point on the y axis, and finding a series of x points along with the change of the y value to determine the initial value of the road boundary;
s4, determining an effective road boundary: and obtaining the effective road boundary based on x and y coordinates by performing smooth analysis and correction on the initial value of the road boundary.
Preferably, in step S1, the scanning probe is disposed on an unmanned aerial vehicle, and the manner of acquiring the field telemetry data specifically includes: the unmanned aerial vehicle vertically flies above a relevant road according to a design track, and the scanning probe transversely scans the ground to obtain distance information between the scanning probe on the unmanned aerial vehicle and a ground scanning point.
Preferably, the manner adopted by the lateral scanning includes: laser scanning, ultrasonic scanning.
Preferably, the recognition algorithm takes the x point combination on the horizontal axis corresponding to the vertical axis y as input, and finds the height difference Δ Z mutation component through neuron transformation and weight combination in different levels to obtain an output boundary sequence rb (y), which is specifically expressed as follows:
Figure BDA0002730245900000021
in equation (1-1), the road boundary function f { RB }lr(y) is expressed as a set of matrices, described as the left boundary C following the variation of the ordinate y valuel(y)RBl(y) and the right boundary Cr(y)RBr(y) a set of x points on the abscissa; RB (y) represents boundary points x corresponding to y, C represents an intelligent smooth correction factor, T represents a transposed matrix, and C is generated by a rule inference set of an artificial intelligence system according to adjacent boundary points and boundary extension trends in an inference mode;
Figure BDA0002730245900000022
Figure BDA0002730245900000031
the formula (2-1) shows that the road left boundary initial value is formed by the x maximum value of the sudden height difference trend, and the formula (2-2) shows that the road right boundary initial value is formed by the x minimum value of the sudden height difference trend; where rb (y) denotes a boundary point x corresponding to y, x (y) denotes a value of x corresponding to y, σ (Σ W Δ Z) is an artificial neural network inference function, W is an input weight factor of a neuron, Δ Z is a high-difference component parameter in the vicinity of the respective x and y coordinate points, and the output of Δ Z is inferred from 25 neurons in the vicinity of the relevant coordinate point.
Preferably, the recognition algorithm is placed in a cloud center, and can be used for simultaneously calculating multiple sections of roads.
The invention has the beneficial effects that: the invention adopts the artificial intelligent neural network technology to calculate, analyze and judge the field telemetering data acquired by the unmanned aerial vehicle through distance measuring methods such as laser, ultrasound or other radio waves, realizes the purpose of automatically giving the boundary of road measurement, and provides scientific basis for determining the planeness and the area of the road surface and the calculation range of other parameters. The invention can analyze the boundary of the road surface without the kerbstone, thereby realizing the rapid, accurate and effective judgment of the road surface range, providing an effective boundary for the further analysis and calculation of the road surface evenness and the road surface effective area, improving the working benefit and the construction quality, saving the human resource, shortening the measurement and calculation time, being beneficial to the operation, the renovation and the acceptance of the relevant road and being beneficial to the scientific construction and the maintenance of the road. The invention can meet the use requirements of the integrated management and the distributed operation of the Internet of things.
The features and advantages of the present invention will be described in detail by embodiments in conjunction with the accompanying drawings.
[ description of the drawings ]
FIG. 1 is a schematic diagram of a system architecture for a road boundary telemetry recognition algorithm using artificial neural network technology according to the present invention;
FIG. 2 is a schematic block diagram of three-dimensional coordinates of a road boundary telemetry recognition algorithm using artificial neural network technology in accordance with the present invention;
FIG. 3 is a schematic diagram of a left boundary artificial neural network model structure of a road boundary telemetry recognition algorithm using an artificial neural network technology according to the present invention.
[ detailed description ] embodiments
Fig. 1 is a schematic structural diagram of a system adopted by the invention, which mainly comprises a road boundary telemetering identification algorithm arranged in a cloud center and a data telemetering system working on site.
Referring to fig. 2, fig. 3 illustrates a road boundary telemetry recognition algorithm using an artificial neural network technique according to the present invention as a calculation method for calculating, reasoning, and analyzing field telemetry data.
The field remote data is that the unmanned aerial vehicle flies vertically above a relevant road according to a designed track, the scanning probe scans the ground transversely through laser, ultrasonic waves or other radio waves (the probe swings on the same road track point and scans the ground), and the acquired distance information between the unmanned aerial vehicle and the ground scanning point is obtained.
The calculation method comprises the steps of firstly setting a three-dimensional coordinate system (as shown in figure 2), defining a longitudinal track as a y axis (on the same y value point, the probe swings to transversely scan the ground, after transverse scanning of one y value is completed, the unmanned aerial vehicle drives the probe to enter the next y value point, the probe swings to scan the next point), and defining transverse scanning as an x axis and defining the height difference between scanning points as a z axis; performing reasoning calculation on adjacent dot matrixes of different x-ray points of the same point on the y axis; according to the change trend of the height difference of adjacent points on the z axis, obtaining the corresponding x point position closest to the boundary at the position where the mutation is generated; along with the change of the y value, a series of x points can be found, so that an initial value of the road boundary is found; and obtaining the effective road boundary based on x and y coordinates by performing smooth analysis and correction on the initial value of the road boundary. Because the calculation method adopts an artificial intelligent neural network method to carry out reasoning calculation, all the calculation methods are called boundary identification algorithms based on artificial neural networks.
The algorithm is placed in the cloud center, and can be used for simultaneously calculating multiple roads. The specific expression of the algorithm is as follows:
Figure BDA0002730245900000041
in equation (1-1), the road boundary function f { RB }lr(y) is expressed as a set of matrices, described as the left boundary C following the variation of the ordinate y valuel(y)RBl(y) and the right boundary Cr(y)RBr(y) set of x points on abscissa. RB (y) denotes the boundary points x corresponding to y, C denotes the Smart smooth correctionThe factor, T, represents the transpose matrix. And C is generated by a rule inference set of a special artificial intelligence system according to adjacent boundary points and boundary extension trend inference.
Figure BDA0002730245900000051
Figure BDA0002730245900000052
The expression (2-1) shows that the road left boundary initial value is formed by the maximum value x of the sudden altitude gradient trend.
And (2-2) the initial value of the right boundary of the road is formed by the minimum value x of the sudden high-difference trend.
Because the y axis of the three-dimensional coordinate axis is arranged in the central part of the road, the x values of the left part of the road are negative values, and the rightmost end of the area generating the altitude difference mutation, namely the corresponding maximum x value, is determined as the initial value of the left boundary; conversely, the x values at the right part of the y axis are all positive values, and the leftmost end of the region generating the step-change, i.e. the corresponding minimum x value, is determined as the initial value of the right boundary.
Where rb (y) denotes a boundary point x corresponding to y, x (y) denotes a value of x corresponding to y, σ (Σ W Δ Z) is an artificial neural network inference function, W is an input weight factor of a neuron, and Δ Z is a high-difference component parameter in the vicinity of the corresponding coordinate point x, y. The output of Δ Z is inferentially generated by 25 neurons in the vicinity of the relevant coordinate point. The artificial neural network reasoning algorithm takes an x point combination on a horizontal axis corresponding to a longitudinal axis y as an input, and finds out a height difference delta Z mutation component through neuron transformation and weight combination in different levels to obtain an output boundary sequence RB (y).
The algorithm searches the mutation value of the ground horizontal axis scanning telemetering data through intelligent reasoning to obtain the x and y coordinate sequence of the road boundary. Because the artificial intelligence reasoning analysis method is adopted, the road boundary can be automatically searched, particularly, the method has great benefits for automatically acquiring the road boundary without the kerbstone, and can meet the use requirements of the integrated management and the distributed operation of the Internet of things.
The above embodiments are illustrative of the present invention, and are not intended to limit the present invention, and any simple modifications of the present invention are within the scope of the present invention.

Claims (5)

1. A road boundary telemetering identification algorithm adopting an artificial neural network technology is characterized in that: the identification algorithm is based on an artificial intelligent neural network, and automatically acquires the effective boundary of the road by performing reasoning calculation on field telemetering data, and comprises the following steps:
s1, acquiring field telemetering data: acquiring distance information between a scanning probe and a ground scanning point;
s2, setting a three-dimensional coordinate system: defining a longitudinal track of a scanning probe as a y-axis, defining transverse scanning to the ground as an x-axis and defining the height difference between ground scanning points as a z-axis;
s3, determining an initial value of a road boundary: calculating corresponding x point positions closest to the boundary at positions where sudden changes occur according to the change trend of the difference of adjacent points on the z axis through the adjacent dot matrix reasoning calculation of different x line points of the same point on the y axis, and finding a series of x points along with the change of the y value to determine the initial value of the road boundary;
s4, determining an effective road boundary: and obtaining the effective road boundary based on x and y coordinates by performing smooth analysis and correction on the initial value of the road boundary.
2. The road boundary telemetry recognition algorithm using artificial neural network technology of claim 1, wherein: in step S1, the scanning probe is disposed on an unmanned aerial vehicle, and the on-site telemetry data acquisition mode specifically includes: the unmanned aerial vehicle vertically flies above a relevant road according to a design track, and the scanning probe transversely scans the ground to obtain distance information between the scanning probe on the unmanned aerial vehicle and a ground scanning point.
3. The road boundary telemetry recognition algorithm using artificial neural network technology of claim 2, wherein: the manner adopted by the lateral scanning includes: laser scanning, ultrasonic scanning.
4. The road boundary telemetry recognition algorithm using artificial neural network technology of claim 1, wherein: the identification algorithm takes an x point combination on a horizontal axis corresponding to a longitudinal axis y as input, finds out a height difference delta Z mutation component through neuron transformation and weight combination in different levels, and obtains an output boundary sequence RB (y), which is specifically expressed as follows:
Figure FDA0002730245890000021
in equation (1-1), the road boundary function f { RB }lr(y) is expressed as a set of matrices, described as the left boundary C following the variation of the ordinate y valuel(y)RBl(y) and the right boundary Cr(y)RBr(y) a set of x points on the abscissa; RB (y) represents boundary points x corresponding to y, C represents an intelligent smooth correction factor, T represents a transposed matrix, and C is generated by a rule inference set of an artificial intelligence system according to adjacent boundary points and boundary extension trends in an inference mode;
Figure FDA0002730245890000022
Figure FDA0002730245890000023
the formula (2-1) shows that the road left boundary initial value is formed by the x maximum value of the sudden height difference trend, and the formula (2-2) shows that the road right boundary initial value is formed by the x minimum value of the sudden height difference trend; where rb (y) denotes a boundary point x corresponding to y, x (y) denotes a value of x corresponding to y, σ (Σ W Δ Z) is an artificial neural network inference function, W is an input weight factor of a neuron, Δ Z is a high-difference component parameter in the vicinity of the respective x and y coordinate points, and the output of Δ Z is inferred from 25 neurons in the vicinity of the relevant coordinate point.
5. The road boundary telemetry recognition algorithm using artificial neural network technology of claim 4, wherein: the recognition algorithm is placed in the cloud center, and can be used for simultaneously calculating multiple sections of roads.
CN202011116151.5A 2020-10-19 2020-10-19 Road boundary remote measurement and identification algorithm adopting artificial neural network technology Withdrawn CN112270689A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011116151.5A CN112270689A (en) 2020-10-19 2020-10-19 Road boundary remote measurement and identification algorithm adopting artificial neural network technology

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011116151.5A CN112270689A (en) 2020-10-19 2020-10-19 Road boundary remote measurement and identification algorithm adopting artificial neural network technology

Publications (1)

Publication Number Publication Date
CN112270689A true CN112270689A (en) 2021-01-26

Family

ID=74337262

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011116151.5A Withdrawn CN112270689A (en) 2020-10-19 2020-10-19 Road boundary remote measurement and identification algorithm adopting artificial neural network technology

Country Status (1)

Country Link
CN (1) CN112270689A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113341417A (en) * 2021-06-09 2021-09-03 深圳市九洲电器有限公司 Detection radar-based road obstacle detection method, vehicle and storage medium

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113341417A (en) * 2021-06-09 2021-09-03 深圳市九洲电器有限公司 Detection radar-based road obstacle detection method, vehicle and storage medium
CN113341417B (en) * 2021-06-09 2024-04-19 深圳市九洲电器有限公司 Road surface obstacle detection method based on detection radar, vehicle and storage medium

Similar Documents

Publication Publication Date Title
WO2020192000A1 (en) Livestock and poultry information perception robot based on autonomous navigation, and map building method
CN106714336A (en) Wireless sensor network temperature monitoring method based on improved Kriging algorithm
Wu et al. Building crack identification and total quality management method based on deep learning
CN110544233B (en) Depth image quality evaluation method based on face recognition application
CN103954300A (en) Fiber optic gyroscope temperature drift error compensation method based on optimized least square-support vector machine (LS-SVM)
CN115376031B (en) Road unmanned aerial vehicle routing inspection data processing method based on federal adaptive learning
CN113781585B (en) Online detection method and system for surface defects of additive manufactured parts
CN112149230A (en) Method for predicting comfort deterioration of wind-induced train of strong wind railway
CN112270689A (en) Road boundary remote measurement and identification algorithm adopting artificial neural network technology
CN112541634B (en) Method and device for predicting top-layer oil temperature and discriminating false alarm and storage medium
CN113532439A (en) Synchronous positioning and map building method and device for power transmission line inspection robot
CN115619953A (en) Rugged terrain-oriented mobile robot terrain mapping method and system
CN116542139A (en) Method and device for predicting roughness of liquid jet polishing surface
CN113850304B (en) High-accuracy point cloud data classification segmentation improvement method
CN115359197A (en) Geological curved surface reconstruction method based on spatial autocorrelation neural network
CN115984360A (en) Method and system for calculating length of dry beach based on image processing
Tian et al. An improved method for NURBS surface based on particle swarm optimization BP neural network
CN114170449A (en) Artificial intelligence image recognition device based on degree of depth learning
CN112784785A (en) Multi-sample fitting image sharpening processing method
CN112950689B (en) Three-dimensional characterization method based on information entropy
Yan et al. Water Quality Detection Based on FCN and Embedded System
Chen et al. Internet of Things-Based Agricultural Mechanization Using Neural Network Extreme Learning on Rough Set
CN117171856B (en) Highway railway digital information modeling method based on handheld Lidar
CN111914402B (en) Dynamic topology estimation system and method based on signal characteristics and topology change priori
Shen et al. CAD Fabric Model Defect Detection Based on Improved Yolov5 Based on Self-Attention Mechanism

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
WW01 Invention patent application withdrawn after publication

Application publication date: 20210126

WW01 Invention patent application withdrawn after publication