CN-122023529-A - Traffic infrastructure dynamic baseline self-adaptive updating method based on low-altitude remote sensing

CN122023529ACN 122023529 ACN122023529 ACN 122023529ACN-122023529-A

Abstract

The invention relates to the technical field of high-precision positioning, in particular to a traffic infrastructure dynamic baseline self-adaptive updating method based on low-altitude remote sensing, which comprises the following steps: the method comprises the steps of acquiring a projection relation between a gesture and an image, extracting a mapping generation data set, extracting lane lines to obtain a normal sequence, constructing a sampling group to calculate offset, correcting coordinates to obtain an update set, and writing the update set into a record table to generate a feedback packet. According to the invention, continuous road sections are extracted through direction angle difference analysis, normal directions are generated, automatic discrimination and unified expression of road geometric features are carried out, transverse sampling is constructed by combining central line vector data, and transverse offset is calculated, so that central line position adjustment has quantitative basis, coordinate correction is carried out on an overrun section along the normal direction, endpoint distance constraint is introduced, line segment continuity and space consistency are ensured, an update result is written into a database in a structured form, feedback identification is formed, and dynamic maintenance capability and overall update efficiency of baseline data are improved.

Inventors

ZHANG CHEN
CHEN JIABIN
LU NA
FAN JIANGTAO
ZHANG ZHENGWEI
ZHAO JING

Assignees

西安航空学院

Dates

Publication Date: 20260512
Application Date: 20260204

Claims (10)

1. The traffic infrastructure dynamic baseline self-adaptive updating method based on low-altitude remote sensing is characterized by comprising the following steps of: s1, acquiring positioning attitude parameters, acquiring road center line vector data stored in a traffic infrastructure baseline database, corresponding to the positioning attitude parameters and remote sensing images according to time sequence, establishing a projection conversion relation, extracting a traffic area ground mapping result, and generating a road area ground mapping data set; s2, extracting a lane line segment coordinate point sequence based on the road region ground mapping data set, identifying continuous coordinate paragraphs with a direction angle difference value smaller than a threshold value, calculating an average value, rotating to form a normal, and generating a normal sequence of a qualified section; S3, constructing a transverse sampling point group by combining the normal sequence of the qualified section and the vector data of the central line, comparing the transverse sampling point group with the pixel coordinate set of the image skeleton line to obtain a transverse distance, calculating transverse offset, and generating a central line transverse offset sequence; S4, correcting coordinate points of the super-threshold line segments in the center line transverse offset sequence, comparing the distances between the end point coordinates of the corrected segment and the end points of the uncorrected segment, adjusting the end point coordinates according to the normal direction of the qualified segment when the conditions are not met, and executing coordinate correction again based on the adjusted end point coordinates until the distance between the end points meets the connection conditions or reaches preset correction judgment conditions, so as to generate a center line updating coordinate set; And S5, constructing a structured data item based on the central line updating coordinate set, writing the structured data item into a baseline updating record table, extracting an offset direction and an offset value as state identifiers, establishing an adaptive feedback field, and generating a central line updating feedback data packet.
2. The traffic infrastructure dynamic baseline adaptive updating method based on low-altitude remote sensing according to claim 1, wherein the road area ground mapping data set comprises a flight platform positioning parameter, a flight platform attitude parameter, remote sensing image pixel coordinates and traffic area ground coordinates, the qualified section normal sequence comprises a lane line segment coordinate point sequence, a direction angle difference value, a direction angle average value and a section normal direction, the center line transverse offset sequence comprises center line vector data, transverse sampling point coordinates, image skeleton line pixel coordinates and transverse direction offset, the center line updating coordinate set comprises corrected segment endpoint coordinates, uncorrected segment endpoint coordinates, endpoint distance data and corrected center line points, and the center line updating feedback data packet comprises a structured data item, a baseline updating record, an offset direction identifier, an offset value identifier and an adaptive adjustment feedback field.
3. The traffic infrastructure dynamic baseline self-adaptive updating method based on low-altitude remote sensing as claimed in claim 1, wherein the positioning gesture parameters comprise position parameters, gesture angle parameters and time stamps, and the time window threshold value used for corresponding in time sequence is used for limiting the maximum matching time difference between the positioning gesture parameters and the remote sensing images.
4. The method for dynamically updating the base line of the traffic infrastructure based on low-altitude remote sensing as claimed in claim 1, wherein the direction angle difference value threshold is used for judging the direction consistency of adjacent line segments in the coordinate point sequence of the lane line segments.
5. The traffic infrastructure dynamic baseline adaptive updating method based on low-altitude remote sensing according to claim 1, wherein the offset threshold is used for judging whether the offset of the intersection of the center line exceeds an offset range, and the connection threshold is used for judging connection continuity of the corrected segment end point and the uncorrected segment end point.
6. The traffic infrastructure dynamic baseline adaptive updating method based on low-altitude remote sensing according to claim 1, wherein the specific steps of S1 are as follows: S101, acquiring positioning and attitude parameters of a flight platform, synchronously acquiring remote sensing image data streams and digital elevation reference data covering traffic areas, analyzing timestamp information of images and parameter records, calculating time synchronization deviation values of the images and the parameter records, screening positioning and attitude data with the deviation values within a preset tolerance range, carrying out associated registration on the data and corresponding image frames, and generating a time sequence image set with the attitude parameters; s102, invoking the time sequence image set with the gesture parameters, extracting platform position coordinates and gesture angle data, constructing an external azimuth element matrix, combining digital elevation reference data to execute space coordinate back calculation operation, calculating ground projection coordinate positions corresponding to image pixel points, converting image coordinate system data into ground space coordinate system data, and establishing a full-frame pixel geographic projection mapping table; and S103, identifying pixel point indexes of traffic road network areas in the images based on the full-width pixel geographic projection mapping table, extracting corresponding ground projection coordinate data according to the indexes, marking interference coordinate information of background areas, aggregating spatial positions and texture feature data of pixels of the road areas, and carrying out serialization recombination on discrete data point sets to obtain the road area ground mapping data set.
7. The traffic infrastructure dynamic baseline adaptive updating method based on low-altitude remote sensing according to claim 1, wherein the specific steps of S2 are as follows: S201, based on the road area ground mapping data set, extracting a lane line segment pixel point set, carrying out communication aggregation, carrying out serialization sequencing on the pixel points according to the space trend, completing index binding and paragraph identification on the coordinate point sequence, and obtaining a lane line segment coordinate point sequence; S202, calling the coordinate point sequence of the lane line segment, carrying out difference calculation on the direction angles of the adjacent coordinate points, generating a direction angle difference value sequence, comparing and judging the direction angle difference value sequence with a preset direction angle difference value threshold value, marking index intervals with the direction angle difference value smaller than the threshold value, forming a continuous coordinate segment set according to the index intervals in an aggregation mode, carrying out segment numbering filing on the continuous coordinate segment set, and generating a qualified coordinate segment set; S203, based on the qualified coordinate paragraph set, carrying out mean value calculation on the direction angle difference value in each paragraph to obtain a paragraph direction angle mean value sequence, executing normal rotation transformation according to the paragraph direction angle mean value sequence to generate a paragraph normal vector sequence, sequentially splicing the paragraph normal vector sequences to complete index binding, establishing a paragraph normal association mapping table, and generating a qualified section normal sequence.
8. The traffic infrastructure dynamic baseline adaptive updating method based on low-altitude remote sensing according to claim 1, wherein the specific steps of S3 are as follows: S301, calling the center line vector data and a normal sequence of a qualified section, constructing transverse sampling point groups on two sides of a center line segment according to a normal direction, sequentially generating a normal coordinate extension path for each center line segment, sampling on the paths, and completing numbering binding of all sampling points according to paragraph sequences to obtain a center line transverse sampling point set; S302, according to the center line intersecting sampling point set, space coordinates of sampling points are spatially matched with an image skeleton line pixel coordinate set, the transverse distance between the sampling points and the skeleton line is calculated, and the distance values are classified and bound according to the center line segment index to generate a sampling point transverse distance sequence; S303, calling the transverse distance sequence of the sampling points, carrying out paragraph inner distance value aggregation processing according to the central line segment index group, calculating the transverse direction offset corresponding to each central line segment, and carrying out sequencing coding on all the offsets according to the index sequence to generate a central line transverse offset sequence.
9. The traffic infrastructure dynamic baseline adaptive updating method based on low-altitude remote sensing according to claim 1, wherein the specific steps of S4 are as follows: S401, carrying out coordinate point correction operation on a central line segment marked as exceeding an error threshold value in the central line transverse offset sequence along the normal direction, extracting an original coordinate point sequence for each central line segment, performing offset mapping, finishing the recombination of corrected coordinate point sequences, establishing paragraph index association between a corrected segment and an uncorrected segment, and generating an out-of-tolerance segment corrected coordinate sequence; S402, calling the out-of-tolerance segment correction coordinate sequence, extracting correction segment end point coordinates and adjacent uncorrected segment end point coordinates, calculating Euclidean distance between end points, comparing the Euclidean distance with a preset end point connection distance threshold value, identifying end point pairs with the distance value smaller than the end point connection distance threshold value, recording corresponding coordinate positions, classifying numbers, and obtaining an end point distance judgment set; s403, based on the endpoint distance judging set, calling endpoint coordinate number information to execute corresponding endpoint replacement operation on the out-of-tolerance segment correction coordinate sequence, reconstructing the correction segment coordinate point sequence and updating the segment structure, and executing continuity retrieval and space consistency identification on the segment connection state by combining the replaced result to generate a center line update coordinate set.
10. The traffic infrastructure dynamic baseline adaptive updating method based on low-altitude remote sensing according to claim 1, wherein the specific steps of S5 are as follows: S501, calling the central line updating coordinate set, carrying out structural organization on the coordinate point sequence of each central line segment and the corresponding paragraph index, establishing a field mapping table according to the road number, the paragraph position and the coordinate attribute, extracting coordinate attribute fields of all central line segments, writing the coordinate attribute fields into a baseline updating record table of a traffic infrastructure database, and generating a central line structural record table; S502, based on the central line structured record table, extracting a transverse offset direction and an offset value corresponding to each central line segment, carrying out direction coding processing on an offset direction field, carrying out amplitude normalization processing on the offset value field, binding a processing result into a state identification field group, and obtaining a central line offset state identification set; S503, calling the center line offset state identification set, writing the offset direction and the offset value into a database field extension configuration set as feedback control parameters, executing binding operation according to the state identification field mapping field set, aggregating the structured record table and the feedback field content to generate a callbacks data object, and generating a center line updating feedback data packet.

Description

Traffic infrastructure dynamic baseline self-adaptive updating method based on low-altitude remote sensing Technical Field The invention relates to the technical field of high-precision positioning, in particular to a traffic infrastructure dynamic baseline self-adaptive updating method based on low-altitude remote sensing. Background The method mainly comprises the steps of establishing a positioning reference, unifying a coordinate system, acquiring positioning data, calibrating positioning results and controlling errors, acquiring positioning observables through a satellite positioning receiver, completing positioning calculation by combining reference station differential correction information, and carrying out coordinate registration with ground control points by combining attitude parameters provided by an inertial measurement unit to form a high-precision position result, wherein the traditional low-altitude remote sensing-based traffic infrastructure dynamic baseline self-adaptive updating method is characterized in that a low-altitude flight platform is utilized to acquire a traffic infrastructure area remote sensing image, the existing baseline data is taken as a reference, the spatial positions and boundaries of facility elements such as roads and bridges are updated, the coordinate consistency of the traffic infrastructure baseline data under the changing conditions is maintained and the elements are updated, the traditional method adopts an unmanned plane to acquire an orthographic image, combines the airborne satellite positioning data with the image control points, carries out orthographic correction by taking the existing vector such as a road center line, a bridge contour line and the like as a basis, manually drawing and updating, and writing the updating result into a facility geographic data base In the low-altitude remote sensing mode, an orthographic image is acquired by an unmanned aerial vehicle, and an image correction mode is completed by combining airborne satellite positioning data and image control points, so that projection distortion caused by factors such as coordinate errors, attitude disturbance and the like exists in the conversion process between the image and the ground, and the ground accuracy of image interpretation is affected. When the existing baseline data is used as a reference for manual sketching and updating, the road boundary and the geometric form thereof are judged according to the understanding capability and the operation proficiency of operators on the image, so that the efficiency is low, subjective differences exist in the result, and the data consistency is difficult to ensure. The central line and the facility outline are not unified in structural operation flow in the comparison and correction process, vector data are often adjusted only by experience, objective quantitative analysis on the continuity of road forms and local offset is not available, and the problems of boundary jump, line segment splicing fracture and the like after updating are easy to occur. In the process of writing the updating result into the database, an explicit feedback mechanism is not established, so that the self-adaptive adjustment of the future updating behavior cannot be realized, the data maintenance is in a static management state, and the dynamic requirement of facility change is difficult to respond. For example, at a viaduct junction or a ramp region, a significant error is often generated due to view angle shielding or image distortion based on boundary adjustment of manual identification, so that the logical connectivity and space coordination of the whole road network are affected, and the reliability and practical value of traffic infrastructure data are reduced. Disclosure of Invention In order to solve the technical problems in the prior art, the embodiment of the invention provides a traffic infrastructure dynamic baseline self-adaptive updating method based on low-altitude remote sensing, which comprises the following steps: s1, acquiring positioning attitude parameters, acquiring road center line vector data stored in a traffic infrastructure baseline database, corresponding to the positioning attitude parameters and remote sensing images according to time sequence, establishing a projection conversion relation, extracting a traffic area ground mapping result, and generating a road area ground mapping data set; s2, extracting a lane line segment coordinate point sequence based on the road region ground mapping data set, identifying continuous coordinate paragraphs with a direction angle difference value smaller than a threshold value, calculating an average value, rotating to form a normal, and generating a normal sequence of a qualified section; S3, constructing a transverse sampling point group by combining the center line vector data and the normal sequence of the qualified section, comparing the transverse sampling point group with an image skeleton l