阅读说明：本技术 一种铁路无人机视频定位方法 (Video positioning method for railway unmanned aerial vehicle ) 是由 王凯 高文峰 邓继伟 赵海 尹传恒 张英杰 岳亮 高帅 葛玉辉 赵罗明 于 2020-03-26 设计创作，主要内容包括：本发明涉及一种铁路无人机视频定位方法,包括如下步骤：S1建立铁路无人机巡线视频定位基准线位里程-坐标空间索引关系；S2进行巡线视频数据采集以及地理信息编码；S3进行巡线视频数据空间定位,包括构建视频空间定位模型与视频空间定位；S4巡线视频数据里程定位；S5巡线视频与地图场景同步；S6多期次巡线视频同步。本发明将视频定位与铁路线路巡线密切结合,解决了巡线数据快速应用的技术瓶颈；本发明对巡线视频进行汇总和梳理从而实现科学管理；通过构建视频空间定位模型与视频空间定位,将空间、里程位置定位与视频数据建立联系,提高数据的使用效率；本发明还采用巡线视频与地图场景同步与多期次巡线视频同步实现了铁路巡线视频的快速定位。(The invention relates to a video positioning method for a railway unmanned aerial vehicle, which comprises the following steps: s1, establishing a coordinate space index relation between the position mileage of the railway unmanned aerial vehicle line patrol video positioning reference line and the coordinate space index; s2, acquiring line patrol video data and encoding geographic information; s3, performing line patrol video data space positioning, including constructing a video space positioning model and video space positioning; s4, mileage positioning of the inspection video data; s5, synchronizing the line patrol video and the map scene; and S6, performing line patrol video synchronization for multiple periods. The invention closely combines video positioning and railway line patrol, and solves the technical bottleneck of quick application of line patrol data; the invention collects and combs the inspection video so as to realize scientific management; the space and mileage position positioning is linked with the video data by constructing a video space positioning model and video space positioning, so that the use efficiency of the data is improved; the invention also realizes the rapid positioning of the railway line patrol video by adopting the synchronization of the line patrol video and the map scene and the synchronization of the line patrol videos for a plurality of times.)

1. A video positioning method for a railway unmanned aerial vehicle is characterized by comprising the following steps: s1, establishing a coordinate space index relation between the position mileage of the railway unmanned aerial vehicle line patrol video positioning reference line and the coordinate space index; s2, acquiring line patrol video data and encoding geographic information; s3, performing line patrol video data space positioning, including constructing a video space positioning model and video space positioning; s4, mileage positioning of the inspection video data; s5, synchronizing the line patrol video and the map scene; and S6, performing line patrol video synchronization for multiple periods.

2. The railway drone video positioning method of claim 1, characterized in that: the S1 comprises the following steps of taking a vector central line of a railway trunk line as a reference for positioning the unmanned aerial vehicle, discretizing the vector central line according to a certain distance through a vector line, and converting the position of the vector line into discrete points; and converting the vector line position into a coordinate point, and establishing a mileage-coordinate index relation.

3. The railway drone video positioning method of claim 1, characterized in that: in S2, carry out unmanned aerial vehicle flight log data interception according to the start-stop time of unmanned aerial vehicle video data collection, obtain spatial position, flight attitude, the angle of aircraft cloud platform, camera parameter and the data acquisition date of unmanned aerial vehicle video recording in-process and match above information with this moment as the benchmark, acquire unmanned aerial vehicle POS data.

4. The railway drone video positioning method of claim 3, characterized in that: the S3 comprises the following steps of establishing an unmanned aerial vehicle video space positioning model by using unmanned aerial vehicle POS data as a data source, and obtaining video key frame positions, key frame projection center position coordinates and video coverage space range coordinate index information through model calculation; matching the projection center coordinates of the video key frames with the coordinates of the input positioning points to obtain optimal matching video frames; and further confirming the spatial position relation between the spatial point location and the matched video key frame, correspondingly locating the position in the video range, namely finding the position of the video frame, and determining the video playing position and time information.

5. The railway drone video positioning method of claim 4, characterized in that: in the step S4, the located railway mileage position is input, the center line mileage information is converted into the space point location coordinate information of the position through the mileage-coordinate index file established in the step S1, and the video space location is carried out according to the space coordinate location method of the step S3 by utilizing the location point coordinate information.

6. The railway drone video positioning method of claim 4, characterized in that: the S5 comprises the following steps of firstly, carrying out interpolation encryption on video POS data to determine the space position coordinates and posture information of the unmanned aerial vehicle at each frame position; calculating the space point coordinate corresponding to the video projection center and the space position covered by each video frame through the unmanned aerial vehicle video space positioning model established in the S3; carrying out map scene positioning and map scene range determination by using the central coordinates and the corner coordinates, and sequentially playing a video and dynamically calculating the spatial position information of a video frame by taking a time sequence as a reference; and updating the map display range by using the calculated spatial position information of each video frame.

7. The railway drone video positioning method of claim 6, characterized in that: the spatial position covered by each video frame is described in a four-corner coordinate manner.

8. The railway drone video positioning method of claim 4, characterized in that: the S6 comprises the following steps of firstly, respectively carrying out interpolation encryption on two-stage video data according to POS data of the unmanned aerial vehicle, and determining the space position coordinates and the posture information of the unmanned aerial vehicle at each frame position; respectively carrying out point coordinate calculation of central projection points of two-stage video frames by using the unmanned aerial vehicle video space positioning model constructed in the S3; taking the first-stage video data as a reference, performing video frame matching on the second-stage video data by using a Euclidean distance method by using the calculated point position coordinates as positioning point coordinates, acquiring the frame matching time position of the second-stage video data, and performing skip playing on the second-stage video; and taking time as a sequence, synchronously playing the videos in two phases, dynamically calculating the Euclidean distance between the central projection coordinates of the positions of the video frames, continuously playing if the error limit requirement is met, and repositioning and matching the videos exceeding the error limit by taking the video in the first phase as a reference.

9. The method of claim 7 or 8, wherein: the interpolation encryption step is specifically to perform interpolation encryption on the key frame POS data of the unmanned aerial vehicle by using a linear interpolation method on the assumption that the change of the position and the posture of the unmanned aerial vehicle is uniform along with time in a short time.

Technical Field

The invention belongs to the technical field of unmanned aerial vehicles, and relates to a video positioning method for a railway unmanned aerial vehicle, in particular to a method for carrying out railway line patrol video space positioning, railway mileage positioning, video scene synchronization and multi-phase video synchronization along a railway by using an unmanned aerial vehicle platform.

Background

With the rapid development of railway construction in China, the railway mileage and the coverage area are continuously increased, and the railway operation mileage in China exceeds 13.9 kilometers by the end of 2019. In order to ensure the normal operation of a railway trunk line, 100 ranges of two sides of the railway line are set as railway surrounding environment protection areas, fence damage exists in the protection areas for a long time, and phenomena such as illegal buildings, industrial dumping, domestic garbage and the like are built in the protection areas. The existence of these phenomena makes the operation of railway receive very big threat, and the traditional mode is patrolled the line through the manual work, carries out the peripheral environment of railway and patrols, and in recent years, along with the rapid development of unmanned aerial vehicle technique, traditional manual work is patrolled the line and is gradually replaced. However, the unmanned aerial vehicle patrols the line video a lot of volume, can't accomplish scientific management and space, mileage position location, has seriously influenced the availability factor of data. The research is developed aiming at the current situation, and the rapid positioning of the railway line patrol video is realized through technical methods such as unmanned aerial vehicle flight position, attitude restoration and video matching.

Disclosure of Invention

The invention provides a railway unmanned aerial vehicle video positioning method for solving the technical problems in the known technology, and aims to realize the quick positioning of the railway unmanned aerial vehicle line patrol video data.

The invention comprises the following technical scheme: a video positioning method for a railway unmanned aerial vehicle comprises the following steps: s1, establishing a coordinate space index relation between the position mileage of the railway unmanned aerial vehicle line patrol video positioning reference line and the coordinate space index; s2, acquiring line patrol video data and encoding geographic information; s3, performing line patrol video data space positioning, including constructing a video space positioning model and video space positioning; s4, mileage positioning of the inspection video data; s5, synchronizing the line patrol video and the map scene; and S6, performing line patrol video synchronization for multiple periods.

The S1 comprises the following steps of taking a vector central line of a railway trunk line as a reference for positioning the unmanned aerial vehicle, discretizing the vector central line according to a certain distance through a vector line, and converting the position of the vector line into discrete points; converting the vector line position into a coordinate point, and establishing a mileage-coordinate index relation; and constructing corresponding scene index data of the line patrol work point-mileage-coordinate according to the line patrol work point odometer and the mileage-coordinate index relation.

In S2, carry out unmanned aerial vehicle flight log data interception according to the start-stop time of unmanned aerial vehicle video data collection, obtain spatial position, flight attitude, the angle of aircraft cloud platform, camera parameter and the data acquisition date of unmanned aerial vehicle video recording in-process and match above information with this moment as the benchmark, acquire unmanned aerial vehicle POS data.

In order to obtain a better ground surface stereo effect, in the video recording process of the unmanned aerial vehicle, the flying height h of the unmanned aerial vehicle relative to the ground surface is 100m ═ h ═ 150m, and the angle α of the aircraft holder is 25 degrees ═ α ═ 30 degrees.

The S3 comprises the following steps of establishing an unmanned aerial vehicle video space positioning model by using unmanned aerial vehicle POS data, and obtaining video key frame position, key frame projection center position coordinates and video coverage space range coordinate index information through model calculation; matching the projection center coordinates of the video key frames with the coordinates of the input positioning points to obtain optimal matching video frames; and further confirming the spatial position relation between the spatial point location and the matched video key frame, correspondingly locating the position in the video range, namely finding the position of the video frame, and determining the video playing position and time information.

In the step S4, the located railway mileage position is input, the center line mileage information is converted into the space point location coordinate information of the position through the mileage-coordinate index file established in the step S1, and the video space location is carried out according to the space coordinate location method of the step S3 by utilizing the location point coordinate information.

The S5 comprises the following steps of firstly, carrying out interpolation encryption on video POS data to determine the space position coordinates and posture information of the unmanned aerial vehicle at each frame position; calculating the space point coordinate corresponding to the video projection center and the space position covered by each video frame through the unmanned aerial vehicle video space positioning model established in the S3; carrying out map scene positioning and map scene range determination by using the central coordinates and the corner coordinates, and sequentially playing a video and dynamically calculating the spatial position information of a video frame by taking a time sequence as a reference; and updating the map display range by using the calculated spatial position information of each video frame.

The spatial position covered by each video frame is described in a four-corner-point coordinate mode, so that the scene positioning and the field range determination are facilitated.

The S6 comprises the following steps of firstly, respectively carrying out interpolation encryption on two-stage video data according to POS data of the unmanned aerial vehicle, and determining the space position coordinates and the posture information of the unmanned aerial vehicle at each frame position; respectively carrying out point coordinate calculation of central projection points of two-stage video frames by using the unmanned aerial vehicle video space positioning model constructed in the S3; taking the first-stage video data as a reference, performing video frame matching on the second-stage video data by using a Euclidean distance method by using the calculated point position coordinates as positioning point coordinates, acquiring the frame matching time position of the second-stage video data, and performing skip playing on the second-stage video; and taking time as a sequence, synchronously playing the videos in two phases, dynamically calculating the Euclidean distance between the central projection coordinates of the positions of the video frames, continuously playing if the error limit requirement is met, and repositioning and matching the videos exceeding the error limit by taking the video in the first phase as a reference.

The interpolation encryption step is specifically to perform interpolation encryption on the key frame POS data of the unmanned aerial vehicle by using a linear interpolation method on the assumption that the change of the position and the posture of the unmanned aerial vehicle is uniform along with time in a short time.

The invention has the advantages and positive effects that:

1. the invention closely combines video positioning and railway line patrol, and solves the technical bottleneck of quick application of line patrol data.

2. According to the invention, scientific management is realized by establishing a railway unmanned aerial vehicle line patrol video positioning benchmark line position mileage-coordinate space index relation, collecting line patrol video data and coding geographic information to collect and comb the unmanned aerial vehicle line patrol video.

3. According to the invention, the space and mileage position location is linked with the video data by constructing the video space location model and the video space location, so that the use efficiency of the data is improved.

4. The invention also realizes the rapid positioning of the railway line patrol video by adopting the synchronization of the line patrol video and the map scene and the synchronization of the line patrol videos for a plurality of times.

5. The unmanned aerial vehicle video positioning method realizes geographic informatization of video multimedia data, converts the video multimedia data into spatial data with geographic reference, and provides a better theoretical basis for spatial big data analysis of video data.

Drawings

FIG. 1 is a video spatial localization method architecture diagram;

FIG. 2 is a video odometry positioning method architecture diagram;

FIG. 3 is a video scene synchronization method architecture diagram;

FIG. 4 is a diagram of a multi-phase video synchronization method architecture;

FIG. 5 is a scene, video synchronization presentation;

FIG. 6 is a two-phase video comparison;

FIG. 7 is a schematic view of vector line bit discretization;

FIG. 8 is a geographic coordinate-spatial coordinate-mileage translation;

FIG. 9 is a geographic coordinate-spatial coordinate-mileage translation-work point translation;

FIG. 10 is a schematic line shot;

fig. 11 is a geographical information encoded file (key frame POS data).

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.