阅读说明：本技术 一种基于点云和影像融合技术的铁路营业线三维中线制作方法 (Railway business line three-dimensional center line manufacturing method based on point cloud and image fusion technology ) 是由 邓继伟 刘成 高文峰 赵海 张冠军 聂虎啸 周文明 王凯 高帅 岳亮 尹传恒 于 2020-03-26 设计创作，主要内容包括：本发明涉及一种基于点云和影像融合技术的铁路营业线三维中线制作方法,包括以下步骤,S1外业控制测量；S2航摄数据获取；S3数据预处理；S4激光点云与影像密集点云配准融合；S5点云切割；S6铁路营业线实景三维重建；S7轨道中心特征点提取；S8铁路线路中心点特征提取；S9铁路三维中线制作。本发明采用激光点云与影像密集点云配准融合,有效提升了激光雷达点云的密度、完整性和平滑性；本发明进行了铁路营业线实景三维重建,该实景三维场景既包含了高精度的几何特征,又包含了丰富的纹理属性,便于量测和判别；本发明有效减少传统铁路营业线测绘的工作量,并大幅提高了外业作业安全性,具有很大的实际应用价值。(The invention relates to a method for manufacturing a three-dimensional center line of a railway business line based on a point cloud and image fusion technology, which comprises the following steps of S1 field control measurement; s2 acquiring aerial data; s3, preprocessing data; s4 registration and fusion of the laser point cloud and the image dense point cloud; s5 point cloud cutting; s6 railway business line live-action three-dimensional reconstruction; s7 extracting characteristic points of the center of the orbit; s8, extracting characteristics of the central point of the railway line; and S9 railway three-dimensional center line manufacturing. According to the invention, the registration and fusion of the laser point cloud and the image dense point cloud are adopted, so that the density, integrity and smoothness of the laser radar point cloud are effectively improved; the invention carries out the live-action three-dimensional reconstruction of the railway business line, and the live-action three-dimensional scene not only contains high-precision geometric characteristics, but also contains rich texture attributes, thereby being convenient for measurement and discrimination; the invention effectively reduces the workload of the traditional railway business line surveying and mapping, greatly improves the field operation safety and has great practical application value.)

1. A railway business line three-dimensional center line manufacturing method based on a point cloud and image fusion technology is characterized by comprising the following steps: s1 field control measurement; s2 acquiring aerial data; s3, preprocessing data; s4 registration and fusion of the laser point cloud and the image dense point cloud; s5 point cloud cutting; s6 railway business line live-action three-dimensional reconstruction; s7 extracting characteristic points of the center of the orbit; s8, extracting characteristics of the central point of the railway line; and S9 railway three-dimensional center line manufacturing.

2. The method for making the three-dimensional center line of the railway business line based on the point cloud and image fusion technology according to claim 1, wherein the method comprises the following steps: and S1 comprises S1-1 of laying plane control point marks along two sides of the railway line position according to rules in the research area and acquiring plane coordinates of the mark points and S1-2 of laying elevation control point marks along two sides of the railway line position according to rules in the research area and acquiring elevation coordinates.

3. The method for making the three-dimensional center line of the railway business line based on the point cloud and image fusion technology of claim 2, wherein: the S2 includes S2-1 obtaining high density laser point cloud of the research area and S2-2 obtaining high resolution digital image of the research area.

4. The method for making the three-dimensional center line of the railway business line based on the point cloud and image fusion technology of claim 3, wherein: the S3 comprises S3-1 laser point cloud data preprocessing and S3-2 image data preprocessing; the step S3-1 is to specifically perform processing on the laser point cloud obtained in the step S2, perform plane and elevation error correction on the laser point cloud by using the control point coordinates obtained in the step S1, and then perform filtering processing on the corrected laser point cloud to remove noise points; the step S3-2 is to process the digital image obtained in the step S2, perform space-triplet encryption and pixel-by-pixel dense matching using the control point coordinates obtained in the step S1 to assist the image, and perform filtering processing based on the image point cloud data obtained by the image dense matching to remove noise points.

5. The method for making the three-dimensional center line of the railway business line based on the point cloud and image fusion technology of claim 4, wherein: and S4, specifically, performing point cloud accurate registration based on the laser point cloud and the image dense point cloud data in the step S3 to obtain fused point cloud data.

6. The method for making the three-dimensional center line of the railway business line based on the point cloud and image fusion technology of claim 5, wherein: and S5, specifically, based on the fused point cloud data obtained in the step S4, buffer areas of 50m on the left and right are manufactured by utilizing the vector line positions of the railway business line, and point clouds are cut to obtain effective point cloud data for mapping the railway business line.

7. The method for making the three-dimensional center line of the railway business line based on the point cloud and image fusion technology of claim 6, wherein: and S6, specifically, performing live-action three-dimensional scene reconstruction based on the air-to-three encryption result obtained in the step S3 and the effective point cloud data obtained in the step S5.

8. The method for making the three-dimensional center line of the railway business line based on the point cloud and image fusion technology of claim 7, wherein: s7 comprises the following steps that S7-1 respectively picks up seed points on the double tracks based on the live-action three-dimensional model obtained in the step S6, and a track surface is extracted by using a region growing algorithm; s7-2, performing edge detection, thinning and line fitting on the rail surface, and dispersing a feature point every 2 meters after removing the rough difference point; s7-3, collecting the field control point coordinates in the step S1 on the live-action three-dimensional model, carrying out subsection adjustment processing by using the field control point coordinates, and carrying out secondary correction on the discrete characteristic points of the track to obtain the corrected accurate coordinates of the characteristic points of the double-track.

9. The method for making the three-dimensional center line of the railway business line based on the point cloud and image fusion technology of claim 8, wherein: s8 specifically comprises the following steps that S8-1 selects coordinates of characteristic points of a left rail track to perform calculation analysis based on the result of the step S7, segments of straight line segments and curve segments are performed on a line according to slope changes of connecting lines of adjacent characteristic points and planes, and inner rail and outer rail identification is performed on double rails of the curve segments; s8-2, obtaining a three-dimensional coordinate of a characteristic point of a central line of the line by interpolating a triangular network formed by two track characteristic points to obtain a central point coordinate; and in the curve section, the characteristic points of the inner rail are shifted to the direction of the central line of the railway by a fixed distance along a plane normal line, and the three-dimensional coordinates of the characteristic points on the central line of the railway are obtained.

10. The method for making the three-dimensional center line of the railway business line based on the point cloud and image fusion technology of claim 9, wherein: and S9, specifically, based on the coordinates of the central point of the railway line obtained in S8 and the existing ledger vector line positions of the railway, gross error elimination and error correction are performed on the central point of the discrete line obtained in the step S8, and finally, the three-dimensional central line of the railway is manufactured through straight line and curve fitting.

Technical Field

The invention belongs to the technical field of railway engineering survey and operation maintenance, and particularly relates to a method for manufacturing a three-dimensional center line of a railway business line based on a point cloud and image fusion technology.

Background

With the development of science and technology, the construction of Chinese railways is changed day by day, the railway mileage and the coverage area are continuously increased, the operating mileage of China railway exceeds 13 kilometers by the end of 2019, wherein the high-speed rail is 3.5 kilometers, and the line-doubling rate and the electrification rate in the operating railway reach over 50 percent. In order to improve the railway traffic and guarantee the railway operation safety, the railway department needs to carry out the works of existing line retesting, electrification transformation, energy expansion and two line increase and the like on partial railways every year, the projects need to carry out three-pass measurement on railway business lines, and the measurement of the railway three-dimensional center line is an important basis of the works.

In order to ensure the normal operation of railways, not only are good lines and related facilities required, but also the requirements on the position precision of the settings are continuously increased, and the geometric state of line operation is particularly important and becomes one of the key factors restricting the capacity expansion or speed increase of railways. The traditional business line measurement adopts pure manual modes such as a steel ruler, a zipper and the like, the efficiency is extremely low, and the precision is poor; with the progress of the surveying and mapping technology, the GPS-RTK means is gradually applied, and although the pure operation efficiency is greatly improved, the time of a skylight of a business line is limited, and the problems of poor safety and difficult on-line are still existed. In recent years, with the rapid development of the unmanned aerial vehicle technology, the unmanned aerial vehicle aerial photography system has the characteristics of simplicity in operation and control, flexibility in use and high cost performance, the application advantages of the unmanned aerial vehicle aerial photography system in the surveying and mapping industry and the railway industry are gradually highlighted, the application range is continuously widened, the three-dimensional center line measurement of the railway business line based on the unmanned aerial vehicle aerial photography system is possible, and at present, no forming technology and method exist.

Disclosure of Invention

The invention provides a method for manufacturing a three-dimensional center line of a railway business line based on a point cloud and image fusion technology, aiming at solving the technical problems in the prior art, the method can be used for efficiently and accurately manufacturing the three-dimensional center line of the line position, providing basic data for subsequent line center coordinate acquisition, plane fitting and plane surveying and mapping, greatly reducing the field work load and effectively improving the safety of railway business line measurement.

The invention comprises the following technical scheme:

a railway business line three-dimensional center line manufacturing method based on point cloud and image fusion technology comprises the following steps: s1 field control measurement; s2 acquiring aerial data; s3, preprocessing data; s4 registration and fusion of the laser point cloud and the image dense point cloud; s5 point cloud cutting; s6 railway business line live-action three-dimensional reconstruction; s7 extracting characteristic points of the center of the orbit; s8, extracting characteristics of the central point of the railway line; and S9 railway three-dimensional center line manufacturing.

And S1 comprises S1-1 of laying plane control point marks along two sides of the railway line position according to rules in the research area and acquiring plane coordinates of the mark points and S1-2 of laying elevation control point marks along two sides of the railway line position according to rules in the research area and acquiring elevation coordinates.

The S2 comprises S2-1, acquiring high-density laser point cloud of a research area and S2-2 high-resolution digital images; the S3 comprises S3-1 laser point cloud data preprocessing and S3-2 image data preprocessing;

the step S3-1 is to specifically perform processing on the laser point cloud obtained in the step S2, perform plane and elevation error correction on the laser point cloud by using the control point coordinates obtained in the step S1, and then perform filtering processing on the corrected laser point cloud to remove noise points; the step S3-2 is to specifically process the image obtained in the step S2, and perform space-three encryption and pixel-by-pixel dense matching by using the control point coordinates obtained in the step S1 to assist the image; filtering the image point cloud data obtained based on image dense matching to remove noise points;

s4, specifically, performing point cloud accurate registration based on the laser point cloud and the image dense point cloud data in the step S3 to obtain fused point cloud data;

s5 is specifically based on the fused point cloud data obtained in the step S4, buffer areas of 50m on the left and right are manufactured by utilizing vector line positions of the railway business line, and point clouds are cut to obtain effective point cloud data for mapping the railway business line;

and S6, specifically, performing live-action three-dimensional scene reconstruction based on the air-to-three encryption result obtained in the step S3 and the effective point cloud data obtained in the step S5. The specific operation requires the use of three-dimensional modeling software, such as oblique photogrammetry processing software like Smart 3D.

S7 comprises the following steps that S7-1 respectively picks up seed points on the double tracks based on the live-action three-dimensional model obtained in the step S6, and a track surface is extracted by using a region growing algorithm; s7-2, performing edge detection, thinning and line fitting on the rail surface, and dispersing a feature point every 2 meters after removing the rough difference point; s7-3, collecting the field control point coordinates in the step S1 on the live-action three-dimensional model, carrying out subsection adjustment processing by using the field control point coordinates, and carrying out secondary correction on the discrete characteristic points of the track to obtain the corrected accurate coordinates of the characteristic points of the double-track.

S8 specifically comprises the following steps that S8-1 selects coordinates of characteristic points of a left rail track to perform calculation analysis based on the result of the step S7, segments of straight line segments and curve segments are performed on a line according to slope changes of connecting lines of adjacent characteristic points and planes, and inner rail and outer rail identification is performed on double rails of the curve segments; s8-2, obtaining a three-dimensional coordinate of a characteristic point of a central line of the line by interpolating a triangular network formed by two track characteristic points to obtain a central point coordinate; and in the curve section, the characteristic points of the inner rail are shifted to the direction of the central line of the railway by a fixed distance along a plane normal line, and the three-dimensional coordinates of the characteristic points on the central line of the railway are obtained.

And S9 is specifically based on the railway line central point coordinates obtained in S8, central line adjustment and fitting are carried out by using least squares, gross error removal and error correction are carried out on the discrete line central points obtained in the step S8 by combining the existing ledger vector line positions of the railway, and finally the three-dimensional central line of the railway is manufactured by fitting straight lines and curves.

The invention has the advantages and positive effects that:

1. according to the invention, the registration and fusion of the laser point cloud and the image dense point cloud are adopted, so that the density, integrity and smoothness of the laser radar point cloud are effectively improved.

2. The invention carries out the live-action three-dimensional reconstruction of the railway business line, and the live-action three-dimensional scene not only contains high-precision geometric characteristics, but also contains rich texture attributes, thereby being convenient for measurement and discrimination.

3. The invention effectively reduces the workload of the traditional railway business line surveying and mapping, greatly improves the field operation safety and has great practical application value.

4. The invention adopts the operation mode of low-altitude photogrammetry and control targets arranged outside the business line, effectively solves the problems of short time and difficult online of a business line skylight, and greatly improves the field work efficiency.

5. The invention uses the multi-source data fusion technology, effectively solves the problems of insufficient information characteristics and coverage of a single data source, and improves the reliability and application value of data.

Drawings

FIG. 1 is a block flow diagram of the present invention.

Fig. 2 is a schematic diagram of track identification.

In the figure, 1-left rail; 2-right rail; 3-outer rail; 4-inner rail.

Detailed Description

To further clarify the disclosure of the present invention, its features and advantages, reference is made to the following examples taken in conjunction with the accompanying drawings.