CN111649752B - Map data processing method, device and equipment for congested road section - Google Patents
Map data processing method, device and equipment for congested road section Download PDFInfo
- Publication number
- CN111649752B CN111649752B CN202010475890.7A CN202010475890A CN111649752B CN 111649752 B CN111649752 B CN 111649752B CN 202010475890 A CN202010475890 A CN 202010475890A CN 111649752 B CN111649752 B CN 111649752B
- Authority
- CN
- China
- Prior art keywords
- image
- point cloud
- point
- time
- track
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
- G01C21/26—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 specially adapted for navigation in a road network
- G01C21/28—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 specially adapted for navigation in a road network with correlation of data from several navigational instruments
- G01C21/30—Map- or contour-matching
- G01C21/32—Structuring or formatting of map data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- 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
- G06T2207/10028—Range image; Depth image; 3D point clouds
-
- 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
- G06T2207/20212—Image combination
- G06T2207/20221—Image fusion; Image merging
Landscapes
- Engineering & Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Theoretical Computer Science (AREA)
- Automation & Control Theory (AREA)
- Traffic Control Systems (AREA)
- Navigation (AREA)
Abstract
The application provides a map data processing method, a map data processing device and map data processing equipment for congested road sections, wherein the method comprises the following steps: the method comprises the steps of acquiring an image of a congested road section acquired in the daytime and a point cloud of the congested road section acquired at night, grouping point clouds to generate a plurality of candidate sets, selecting an associated candidate set associated with the image from the candidate sets, selecting an associated point cloud associated with the image from the associated candidate set, fusing the image and the associated point cloud associated with the image to generate map data of the congested road section. The method provided by the application reduces the point cloud which can be matched with each image, reduces the influence of the image collected in advance on the image collected in the later period, and improves the registration rate of the image and the point cloud so as to obtain the map data of the congested road section.
Description
Technical Field
The application relates to the technical field of geographic information mapping, in particular to a map data processing method, device and equipment for congested road sections.
Background
The high-precision map making is divided into an interior industry and an exterior industry, wherein the exterior industry refers to a process of continuously acquiring data of an acquisition vehicle in a moving state, and the interior industry refers to a process of processing the continuously acquired data to make high-precision map products.
In the process of carrying out field work by the collection vehicle, road congestion conditions are often encountered. Because the vehicle on the road blocks the acquisition vehicle to send out signals, the point cloud acquisition cannot be realized. Urban roads are characterized by congestion in the daytime and smooth night. And aiming at the road congestion condition, images are collected in the daytime, and point clouds are collected at night. And matching the image and the point cloud according to the distance between the image track point and the point cloud track point to obtain map data of the congested road section.
However, when the data acquisition routes overlap, matching of the image and the point cloud is easy to fail, and map data of a congested road section cannot be obtained.
Disclosure of Invention
The application provides a map data processing method, a map data processing device and map data processing equipment for congested road sections, which are used for solving the technical problem that when a turn-back phenomenon exists in a data acquisition route, map data of the congested road sections cannot be generated by an existing processing method in a matching mode.
In a first aspect, the present application provides a map data processing method for a congested road segment, including:
acquiring an image of a congested road section acquired in the daytime and a point cloud of the congested road section acquired at night;
grouping point clouds to generate a plurality of candidate sets;
selecting an associated candidate set associated with the image from a plurality of candidate sets;
selecting an associated point cloud associated with the image from the associated candidate set;
and fusing the image and the associated point cloud associated with the image to generate map data of the congested road section.
Optionally, the grouping processing is performed on the point cloud, and a plurality of candidate sets are generated, which specifically includes:
acquiring time intervals between adjacent point cloud track points at any two groups of acquisition moments, wherein the point cloud comprises point cloud track points;
and grouping the point clouds according to the time intervals and preset segmentation intervals to generate a plurality of candidate sets.
Optionally, grouping the point clouds according to the time interval and the preset segmentation interval to generate a plurality of candidate sets, specifically including:
dividing the time distribution interval according to the time interval to generate a plurality of track time intervals, wherein the time distribution interval is the distribution interval of the acquisition time of the point cloud track points;
dividing each track time interval according to the segmentation interval to generate a grouping time interval;
and grouping the point clouds according to the grouping time interval to generate a plurality of candidate sets.
Optionally, the time distribution interval is divided according to the time interval to generate a plurality of trajectory time intervals, and the method specifically includes:
if the time interval is larger than a preset interval threshold, determining the partitioning time according to the acquisition time of the point cloud track points corresponding to the time interval;
and dividing the time distribution interval according to the partition time to generate a track time interval.
Optionally, the interval threshold is determined according to the collection frequency of the collection vehicle.
Optionally, grouping the point clouds according to the grouping time interval to generate a plurality of candidate sets, specifically including:
and point clouds corresponding to the point cloud track points with the collection time in the same grouping time interval are distributed to the same candidate set.
Optionally, selecting an associated candidate set associated with the image from the plurality of candidate sets specifically includes:
acquiring a first candidate set where an associated point cloud of an image corresponding to an image acquired at the previous moment is located;
determining a second candidate set which is immediately behind the first candidate set when being sorted according to the acquisition time according to the first candidate set;
and taking the first candidate set and the second candidate set as association candidate sets.
Optionally, selecting an associated point cloud associated with the image from the associated candidate set specifically includes:
and selecting the point cloud corresponding to the point cloud track point with the minimum acquisition time as the associated point cloud associated with the image corresponding to the image track point at the current time, wherein the distance between the point cloud track point and the image track point at the current time is within a preset range in the associated candidate set.
In a second aspect, the present application provides a map data processing apparatus for a congested road segment, including:
the acquisition module is used for acquiring an image of a congested road section acquired in the daytime and a point cloud of the congested road section acquired at night;
the grouping module is used for grouping the point clouds to generate a plurality of candidate sets;
a determination module for selecting an associated candidate set associated with the image from a plurality of candidate sets;
the determining module is further used for selecting an associated point cloud associated with the image from the associated candidate set;
and the generation module is used for fusing the image and the associated point cloud associated with the image to generate map data of the congested road section.
In a third aspect, the present application provides a processing apparatus comprising:
a memory for storing a program;
a processor for executing the program stored in the memory, the processor being configured to perform the map data processing method for a congested road segment according to the first aspect and the alternative when the program is executed.
The application provides a map data processing method, device and equipment for congested road sections, which are characterized in that point clouds are grouped to generate a plurality of candidate sets, then associated candidate sets associated with images are determined from the candidate sets, associated point clouds associated with the images are selected from the associated candidate sets, the point clouds which can be matched with each image are reduced, the influence of the images acquired in advance on the images acquired in the later stage is reduced, and the registration rate of the images and the point clouds is improved to obtain map data of the congested road sections.
Drawings
FIG. 1 is a diagram illustrating a matching effect in the prior art;
FIG. 2 is a schematic diagram illustrating a map data processing method provided herein;
fig. 3 is a schematic flowchart of a map data processing method according to an embodiment of the present application;
fig. 4 is a schematic storage diagram of data collected during the daytime according to the third embodiment of the present application;
FIG. 5 is a schematic view of a daytime and nighttime collection route provided in the third embodiment of the present application;
fig. 6 is a schematic diagram of data storage acquired at night according to the third embodiment of the present application;
fig. 7 is a schematic distribution diagram of point cloud track points and image track points collected in the third embodiment of the present application;
fig. 8 is an effect diagram of a map data processing method according to a third embodiment of the present application;
FIG. 9 is a schematic diagram of the matching effect of the comparative example provided in the present application;
fig. 10 is a schematic structural diagram of a map data processing apparatus according to a fourth embodiment of the present application;
fig. 11 is a schematic structural diagram of a processing apparatus according to a fifth embodiment of the present application.
Detailed Description
To make the purpose, technical solutions and advantages of the present application clearer, the technical solutions in the present application will be clearly and completely described below with reference to the drawings in the present application, and it is obvious that the described embodiments are some, but not all embodiments of the present application. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
The high-precision map making is divided into field industry and field industry, wherein the field industry refers to the process of collecting map data of vehicles when going out, and the field industry refers to the process of processing the collected map data of the vehicles to make the high-precision map. Wherein, the process of processing the collected data can be divided into: data processing, map drawing, map data format conversion, map data compilation, and high-precision map distribution. When field work is carried out, the collection vehicle continuously collects data on one road, if a congested road section is encountered during field work in the daytime, the data on the road is collected firstly, and then the road is collected once again at night. The method has the advantages that the night traffic flow is small, point clouds with high quality can be acquired, the daytime light is good, images with high quality can be acquired, the images acquired in the daytime are matched with the point clouds acquired at night, map data can be acquired, the map data are used for carrying out subsequent processing, and a high-precision map is prepared.
In the prior art, images acquired in the daytime and point clouds acquired at night are matched according to the distance of track points and the acquisition time, namely, the distance between the image track points and the point cloud track points is determined, and when the distance between the image track points and the point cloud track points is within a preset threshold value, the images corresponding to the image track points and the point cloud track points are matched. However, when there is overlap in the acquisition trace points, a matching failure phenomenon easily occurs.
As shown in fig. 1, 20 sets of cloud track points and 8 sets of image track points need to be matched, point clouds and images are matched according to the principle of the closest distance and the acquisition time sequence, when a 3 rd image track point is matched, the 20 th cloud track point is closer to the 3 rd image track point, the 20 th cloud track point and the 3 rd image track point can be matched, and the 3 rd image track point actually needs to be matched with the 5 th cloud track point, so that a matching error occurs. The acquisition time of the point cloud track points matched with the image track points at the back is larger than the acquisition time of the track points matched with the 3 rd image track, so that the 4 th image track point to the 8 th image track point at the back cannot be matched, and the matching failure occurs.
The application provides a map data processing method, a map data processing device and map data processing equipment for congested road sections, and aims to solve the problems. The inventive concept of the present application is: as shown in fig. 2, in the point clouds collected during the daytime, the point clouds on non-congested road sections can be continuously used, only the point clouds on congested road sections can be collected at night, then the point clouds are matched with the images to obtain map data of congested road sections, in order to improve matching accuracy, the point clouds are grouped, the point clouds and the images are selected to be matched in a grouping result, the number of the point clouds which can be matched with each image is reduced, the images are prevented from being matched with other unrelated point clouds, images collected later at the moment can also be matched with the point clouds, and then the map data of congested road sections are obtained.
Fig. 3 is a flowchart illustrating a map data processing method according to an embodiment of the present application. As shown in fig. 3, a map data processing method provided in the embodiment of the present application includes the following steps:
s101, acquiring an image of a congested road section acquired in the daytime and a point cloud of the congested road section acquired at night.
The urban road congestion is characterized in that congestion is relatively caused in the daytime, and the congestion phenomenon disappears at night. The collection vehicles typically collect map data for congested road segments both day and night.
The map data includes a point cloud and an image. The point cloud comprises point cloud track points and object coordinate data, the object coordinate data can reflect the information of the collected objects, and the point cloud track points comprise position data and collection time of the collection vehicle. The image comprises image track points and pictures, the pictures can reflect collected object information, and the image track points comprise position data and collection time of a collection vehicle.
Because the light is good in daytime, the image quality in the map data is high, and the actual condition of the road can be accurately reflected. And because the number of vehicles at night is small, the collected vehicle emission signals are not easily shielded by the vehicles, the point cloud quality in the map data is high, and the actual condition of the road can be accurately reflected.
And extracting images of the congested road sections from map data acquired in the daytime. And extracting point clouds of the congested road sections from map data of the congested road sections acquired at night. The image comprises image track points, and the point cloud comprises point cloud track points.
And S102, grouping the point clouds to generate a plurality of candidate sets.
The point clouds are grouped according to the point cloud track points to generate a plurality of candidate sets, namely the point clouds corresponding to the point cloud track points adjacent to the acquisition time are grouped into the same candidate set.
S103, selecting an associated candidate set associated with the image from the plurality of candidate sets.
And determining an associated candidate set of the image corresponding to the image track point at the previous moment according to each image of the congested road section, and determining an associated candidate set of the image corresponding to the image track point at the current moment. More specifically, the associated candidate set of the image corresponding to the image track point at the previous time and the candidate set adjacent to the acquisition time are used as the associated candidate set of the image corresponding to the image track point at the current time. And the acquisition time of the candidate set is determined according to the acquisition time of the point cloud track points in the candidate set.
When the image is acquired at the initial moment, the associated candidate set of the image is the candidate set where the point cloud track point corresponding to the earliest point cloud at the acquisition moment is located.
And S104, selecting the associated point cloud associated with the image from the associated candidate set.
And selecting the associated point cloud associated with the image from the associated candidate set according to the distance between the image track point and the point cloud track point and the sequence of the acquisition time.
And S105, fusing the image and the associated point cloud associated with the image to generate map data of the congested road section.
The image track point acquisition time is replaced by the associated point cloud track point acquisition time, so that the fusion of the image and the associated point cloud can be realized, and the map data of the congested road section is obtained.
In the processing method provided by the embodiment of the application, the point clouds are grouped, the point clouds which can be matched with each image are reduced, the image is prevented from being matched with other unrelated point clouds, the image with the previous acquisition time does not influence the matching when the acquisition time is later, and the registration rate of the image and the point clouds is improved.
The following description focuses on a map data processing method for a congested road segment provided in the second embodiment of the present application, where the map data processing method includes the following steps:
s201, acquiring an image of a congestion road section acquired in the daytime and a point cloud of the congestion road section acquired at night.
Here, this step has already been described in detail in S101, and is not described here again.
S202, grouping the point clouds to generate a plurality of candidate sets.
The method comprises the steps of generating a plurality of candidate sets by grouping original point clouds, and determining road section point clouds related to road section images of congested road sections from the candidate sets. The method for generating a plurality of candidate sets by grouping the original point clouds specifically comprises the following steps: and (3) grouping the original point clouds twice, grouping the original point clouds for the first time according to the acquisition intervals between two groups of point cloud track points, grouping the original point clouds for the second time on the basis of the first grouping, and classifying the original point clouds at the acquisition time in the same segmentation interval.
The basic principle of two packets is: and the partitioning time is determined through the acquisition interval and a preset interval threshold, so that the point clouds on the same point cloud track can be partitioned into the same group. And performing more refined grouping on the point clouds in the same track interval, and determining the associated point clouds associated with the image from the grouped candidate set to improve the matching accuracy.
The first grouping specifically comprises: and acquiring a distribution interval of the point cloud, wherein the distribution interval refers to the distribution interval of the acquisition time of the point cloud track points. And traversing original point clouds of the congested road sections acquired at night, and determining the time interval between adjacent point cloud track points at two acquisition times. Determining a preset interval threshold according to the data acquisition frequency of the acquisition vehicle, if the time interval is larger than the preset interval threshold, indicating that the two point cloud track points are two track points on different tracks, determining a partitioning time according to the acquisition time of the two point cloud track points corresponding to the time interval, and partitioning the distribution partition according to the partitioning time to generate a plurality of track intervals.
The second grouping specifically comprises: for each track section, a plurality of packet sections are generated by dividing each track section at a segment interval. And distributing the original point clouds corresponding to the original point cloud track points in the same grouping interval at the acquisition time to the same candidate set.
S203, selecting an associated candidate set associated with the video from the plurality of candidate sets.
Determining road section point clouds associated with road section images corresponding to the road section image track points at the moment k, determining a candidate set where the road section point clouds associated with the road section images corresponding to the road section image track points at the moment k are located, and defining the candidate set as a first candidate set. Wherein k is a positive integer.
And sequencing the candidate sets according to the acquisition time of the point cloud track points, and defining the candidate set with the acquisition time adjacent to the first candidate set as a second candidate set. And taking the first candidate set and the second candidate set as association candidate sets.
S204, selecting the associated point cloud associated with the image from the associated candidate set.
And selecting the distance between the road section image track points at the moment k +1 from the first candidate set and the second candidate set to be within a preset range, and acquiring the point cloud corresponding to the point cloud track point at the minimum moment as the associated point cloud associated with the second intermediate image.
And determining the distance between each point cloud track point and the road section image track point acquired at the moment k +1 in the first candidate set and the second candidate set, and taking the point cloud corresponding to the point cloud track point as the candidate point cloud if the distance is within the threshold range. If only one candidate point cloud exists, the candidate point cloud is used as the associated point cloud associated with the second intermediate image, and if a plurality of candidate point clouds exist, the candidate point cloud with the minimum acquisition time is selected as the associated point cloud associated with the second intermediate image. Wherein, the distance threshold value can be determined according to the acquisition position interval of the acquisition vehicle. For example: the distance threshold range may take 15 m.
And S205, fusing the image and the associated point cloud associated with the image to generate map data of the congested road section.
Here, the step is already described in detail in S105, and is not described here again.
In the map data processing method provided by the embodiment of the application, the point clouds are grouped twice to obtain the candidate set, the point clouds which can be matched with each image are reduced, the candidate set corresponding to each image is determined from the candidate set, the point clouds which can be matched with the images are further reduced, the influence of the images which are close to the front of the acquisition time on the images which are close to the rear of the acquisition time is reduced, and the registration rate of the images and the point clouds is improved.
A map data processing method provided in a third embodiment of the present application is described below with reference to an example, where the method includes the following steps:
s301, acquiring an image of a congested road section acquired in the daytime and a point cloud of the congested road section acquired at night.
More specifically, the method comprises the steps of collecting normal operation of a vehicle in the daytime, collecting point clouds and images of the whole road, recording event points of congestion events when congestion occurs in the operation process, wherein the event points comprise: a start position of the congested link and an end position of the congested link. For example: in the database, an event point file data structure of congestion events is shown in table 1 below. And collecting the point clouds of the congested road sections again according to the same collection sequence in the daytime according to the event points marked in the daytime.
Table 1 event point file data structure of congestion event
In table 1, the event sequence number corresponds to an event type, and a mapping table between the event sequence number and the event type is shown in table 2.
Table 2 mapping table of event sequence number and event type
Event (Chinese) | Event (English) | Serial number | Event type |
Road congestion | TrafficJam | 4 | Line events |
The method comprises the steps of recording a driving route and an actual driving lane during collection aiming at the day operation of a collection vehicle, and when a high-speed local line is collected, wherein the driving route is used for showing the sequence of collecting the local line and a ramp of a road. And if the number of lanes is less than or equal to 5, the collection vehicle runs along the lanes close to the right in the middle. When collecting the ramp, from the connecting point of the ramp and the high-speed line, repeatedly collecting the road section of 1km along the high-speed line, and driving along the right lane in the middle when collecting the repeated road section.
The images acquired during the day may be named ". x + time. jpg" in the following format, time retaining only 9 digits of a hour, minute, second and millisecond, for example: 100911352. jpg. The database may store the image track file in a data structure as shown in table 3. The daytime map data collected during the daytime is stored in a storable medium according to the daytime collection project shown in fig. 4. As shown in fig. 4, it can be seen that the video file and the event point file are located in the same directory.
TABLE 3 data Structure of video track File
Serial number | Item | Content providing method and apparatus | Format | Size and breadth | Whether or not there is a |
|
1 | GPS time | GPS time (second day) | | 13chars | Y | |
2 | Name of photograph | Name of photograph | Character | 28chars | Y | |
3 | Beijing time | Beijing time, accurate to | Postdecimal | 3 bits | 12chars | Y |
4 | |
3 degree of Gauss with the | Postdecimal | 3 bits | 12chars | |
5 | |
3 deg. of gauss with | Postdecimal | 3 bits | 12chars | Y |
6 | H-Ell | Elevation (WGS84) | |
9chars | Y |
Wherein, to gathering car night work, according to real-time road conditions, begin to gather after the road is unobstructed night. The collection route is marked in daytime and is a route of the event points of the jammed road sections, and the collection sequence is completely consistent with the daytime. For example: fig. 5 is a schematic diagram of a day and night acquisition route provided in the first embodiment of the present application, and as shown in fig. 5, if the sequence of acquiring congested road segments first in the day is (c), the acquisition is performed in the same sequence at night. The database may store the point cloud trajectory file in a data structure as shown in table 4. Map data collected at night are stored in a storable medium according to a night collection project shown in fig. 6. As shown in fig. 6, a catalog where the point cloud is located may be obtained.
Table 4 data structure of point cloud track file
And S302, grouping the point clouds to generate a plurality of candidate sets.
The arrival time of the multiple groups of point cloud track points is counted, and the distribution interval U0 of the arrival time of the multiple groups of point cloud track points is determined. The first acquisition time interval Δ T1 is any one of a plurality of acquisition time intervals, and the first acquisition time interval Δ T1 is determined according to the acquisition time T1 of the first point cloud trajectory point and the arrival time T2 of the second point cloud trajectory point.
If the first acquisition time interval delta T1 is larger than a preset interval threshold delta, determining a partition time T0 according to the arrival time of the first point cloud track point and the arrival time of the second point cloud track point, directly taking the arrival time of the first point cloud track point and the arrival time of the second point cloud track point as the partition time T0, averaging the arrival time of the first point cloud track point and the arrival time of the second point cloud track point to generate a partition time T0, and selecting any one of the first point cloud track point and the second point cloud track point as the partition time T0.
The preset interval threshold is determined according to the collection frequency of the collection vehicle. Preferably, the predetermined interval threshold δ is greater than the inverse of the acquisition frequency of the acquisition vehicle, i.e., δ > 1/f. Where f denotes the acquisition frequency.
After the partitioning time T0 is acquired, the distribution section U0 is subjected to the segmentation process using the partitioning time T0 to determine a trajectory section U1. The track section U1 is divided into segment intervals Δ t2, and a grouping section U2 is determined. A time point in the trajectory interval U1 can be arbitrarily selected as a grouping time t31, a time point separated by a section separation Δ t2 from the grouping time t31 is used as a next grouping time t32, and two adjacent grouping time points are end points of the corresponding grouping interval. And grouping the multiple groups of point clouds according to the grouping intervals, distributing the point clouds in the same grouping interval at the acquisition time to the same candidate set, and generating multiple candidate sets.
Fig. 7 is a schematic distribution diagram of point cloud track points and image track points collected in the second embodiment of the present application, and as shown in fig. 7, black points represent point cloud track points, elliptical points represent image track points, there are 20 sets of point cloud track points and 8 sets of image track points, a distribution interval U0 of collection time of 20 sets of point cloud track points is determined to be [0, 270] s through statistics, and the 20 sets of point cloud track points are sequentially marked as a 1 st point cloud track point to a 20 th point cloud track point. The acquisition time interval Δ T1 is as follows: 10s, 10s, 10s, 10s, 10s, 10s, 10s, 10s, 40s, 10s, 10s, 10s, 10s, 10s, 10s, 10s, 60s, 10s, 10s, 10s, 10 s. The preset interval threshold is 12s, and it can be known that the acquisition time interval 40s and the acquisition time interval 60s satisfy the condition. The acquisition time interval 40s is determined according to the arrival time 80s of the 9 th point cloud track point and the arrival time 120s of the 10 th group of point cloud track points. The acquisition time interval 60s is determined according to the acquisition time 180s of the 16 th point cloud track point and the arrival time 240s of the 17 th point cloud track point.
And according to the acquisition time 80s of the 9 th point cloud track point and the acquisition time 120s of the 10 th point cloud track point, taking the arrival time 180s of the 16 th point cloud track point and the arrival time 240s of the 17 th point cloud track point as partition times.
The distribution interval U0[0, 270] s is divided by using the partition time, and the following track interval U1 is determined: u1a [0, 80], U1b [120, 180], U1c [240, 270 ]. And determining grouping time according to a preset segmentation interval delta t2, wherein delta t2 is 50s, and segmenting each track section U1 according to the segmentation interval to obtain a plurality of grouping sections. The following grouping interval U2 can be obtained: u2a [0, 50], U2b [50, 80], U2c [120, 170], U2d [170, 180], U2e [240, 270 ]. And then point clouds falling into the same grouping interval are grouped into the same candidate set.
S303, select an associated candidate set associated with the video from the plurality of candidate sets.
After obtaining a plurality of candidate sets, taking the candidate set matched with the image corresponding to the image track point at the previous acquisition time and the candidate set immediately after the acquisition time as associated candidate sets.
S304, selecting the associated point cloud associated with the image from the associated candidate set.
And selecting the point cloud corresponding to the point cloud track point with the minimum acquisition time as the associated point cloud associated with the image corresponding to the image track point at the current time, wherein the distance between the point cloud track point and the image track point at the current time is within a preset range.
With continued reference to fig. 7, the candidate set U2a, the candidate sets U2b, … …, and the candidate set U2e are obtained in order according to the collection time order. The 1 st road section image track point is an initial track point, an associated point cloud track point is selected from the candidate set U2a, the distance between the 3 rd point cloud track point and the 1 st road section image track point meets a preset threshold value, and the 3 rd point cloud track point and the 1 st road section image track point are matched. And selecting the associated point cloud track points from the candidate set U2a and the candidate set U2b according to the 2 nd road segment image track point and the 3 rd road segment image track point. The distance between the 5 th point cloud track point and the 2 nd road section image track point is within a preset range, and the two points are associated. The distances between the 7 th point cloud track point, the 10 th point cloud track point and the 3 rd road section image track point are within a preset range, and the 7 th point cloud track point and the 3 rd road section image track point are correlated at a time before the acquisition moment. And selecting the associated point cloud track points from the 4 th road segment image track point to the 6 th road segment image track point in the candidate set U2b and the candidate set U2 c. And the distances between the 9 th point cloud track point, the 12 th point cloud track point and the 4 th road section image track point are within a preset range, and the 9 th point cloud track point which is ahead of the acquisition moment is associated with the 4 th road section image track point. And the distance between the 5 th road section image track point and the 14 th point cloud track point meets a preset threshold value, and the two points are matched. And the distance between the 6 th road section image track point and the 11 th point cloud track point meets a preset threshold value, and the two points are matched. And so on.
S305, fusing the image of the daytime congestion road section and the point cloud associated with the image of the daytime congestion road section to generate data of the daytime congestion road section.
Here, this step has already been described in detail in S104, and repeated descriptions are omitted. After the link image of the congested link is matched with the point cloud, the map data of the congested link may be stored according to a storage manner as shown in fig. 8.
The effect of the map data processing method provided by the present application is described below with reference to a comparative embodiment, and fig. 9 is a matching effect diagram provided by the comparative embodiment of the present application, and as shown in fig. 9, three candidate sets can be obtained by grouping point clouds according to a trajectory. Matching the point cloud and the image according to the following matching principle: and selecting associated point clouds of which the distances meet preset values from the candidate set matched with the previous image track point and the candidate set immediately behind the acquisition time, wherein the acquisition time of the point cloud track point matched with the previous image track point is less than the acquisition time of the point cloud track point matched with the next image track point. The candidate sets from the 2 nd image track point to the 6 th image track point are the same and are the candidate set U1a and the candidate set U1 b. After the 5 th image track point is matched with the 13 th image track point, the 6 th image track point cannot be matched, and then the follow-up image track points cannot be matched continuously.
The map data processing method provided by this embodiment performs twice grouping on point clouds to obtain candidate sets, reduces point clouds that can be matched with each image, determines a candidate set corresponding to each image from the candidate sets, further reduces the point clouds that can be matched with the images, reduces the influence of images that are collected earlier on images that are collected later on the images, and improves the registration rate of the images and the point clouds.
Fig. 10 is a schematic structural diagram of a map data processing apparatus for providing a congested road segment according to a fourth embodiment of the present invention, and as shown in fig. 10, a map data processing apparatus 400 according to the fourth embodiment of the present invention includes:
the acquisition module 401 is configured to acquire an image of a congested road section acquired in the daytime and a point cloud of the congested road section acquired at night;
a grouping module 402, configured to perform grouping processing on the point clouds to generate a plurality of candidate sets;
a determining module 403, configured to select an associated candidate set associated with the image from a plurality of candidate sets;
the determining module 403 is further configured to select an associated point cloud associated with the image from the associated candidate set;
and a generating module 404, configured to fuse the image and the associated point cloud associated with the image, and generate map data of the congested road segment.
Optionally, the grouping module 402 is specifically configured to:
acquiring time intervals between adjacent point cloud track points at any two groups of acquisition moments, wherein the point cloud comprises point cloud track points;
and grouping the point clouds according to the time intervals and preset segmentation intervals to generate a plurality of candidate sets.
Optionally, the grouping module 402 is specifically configured to:
dividing the time distribution interval according to the time interval to generate a plurality of track time intervals, wherein the time distribution interval is the distribution interval of the acquisition time of the point cloud track points;
dividing each track time interval according to the segmentation interval to generate a grouping time interval;
and grouping the point clouds according to the grouping time interval to generate a plurality of candidate sets.
Optionally, the grouping module 402 is specifically configured to:
if the time interval is larger than a preset interval threshold, determining the partitioning time according to the acquisition time of the point cloud track points corresponding to the time interval;
and dividing the time distribution interval according to the partition time to generate a track time interval.
Optionally, the interval threshold is determined according to the collection frequency of the collection vehicle.
Optionally, the grouping module 402 is specifically configured to:
and point clouds corresponding to the point cloud track points with the collection time in the same grouping time interval are distributed to the same candidate set.
Optionally, the determining module 403 is specifically configured to:
acquiring a first candidate set where an associated point cloud of an image corresponding to an image acquired at the previous moment is located;
determining a second candidate set which is immediately behind the first candidate set when being sorted according to the acquisition time according to the first candidate set;
and taking the first candidate set and the second candidate set as association candidate sets.
Optionally, the determining module 403 is specifically configured to:
and selecting the point cloud corresponding to the point cloud track point with the minimum acquisition time as the associated point cloud associated with the image corresponding to the image track point at the current time, wherein the distance between the point cloud track point and the image track point at the current time is within a preset range in the associated candidate set.
Fig. 11 is a schematic structural diagram of a processing apparatus according to a fifth embodiment of the present application. As shown in fig. 11, the present embodiment provides a processing apparatus 500 including: a transmitter 501, a receiver 502, a memory 503, and a processor 504.
A transmitter 501 for transmitting instructions and data;
a receiver 502 for receiving instructions and data;
a memory 503 for storing computer-executable instructions;
the processor 504 is configured to execute the computer-executable instructions stored in the memory to implement the steps performed by the map data processing method for congested road segments in the above-described embodiments. Specifically, reference may be made to the related description in the foregoing map data processing method embodiment of the congested road segment.
Alternatively, the memory 503 may be separate or integrated with the processor 504.
When the memory 503 is provided separately, the processing device further includes a bus for connecting the memory 503 and the processor 504.
The embodiment of the application also provides a computer-readable storage medium, wherein a computer executing instruction is stored in the computer-readable storage medium, and when the processor executes the computer executing instruction, the map data processing method of the congested road section, which is executed by the processing device, is realized.
Finally, it should be noted that: the above embodiments are only used for illustrating the technical solutions of the present application, and not for limiting the same; although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; and the modifications or the substitutions do not make the essence of the corresponding technical solutions depart from the scope of the technical solutions of the embodiments of the present application.
Claims (10)
1. A map data processing method for a congested road segment, characterized by comprising:
acquiring an image of a congested road section acquired in the daytime and a point cloud of the congested road section acquired at night;
grouping the point clouds to generate a plurality of candidate sets;
selecting an associated candidate set associated with the image from the plurality of candidate sets;
selecting an associated point cloud associated with the image from the associated candidate set;
and fusing the image and the associated point cloud associated with the image to generate map data of the congested road section.
2. The method according to claim 1, wherein grouping the point clouds to generate a plurality of candidate sets comprises:
acquiring a time interval between any two groups of adjacent point cloud track points at the acquisition time, wherein the point cloud comprises point cloud track points;
and grouping the point clouds according to the time interval and a preset segmentation interval to generate a plurality of candidate sets.
3. The method of claim 2, wherein grouping the point clouds according to a time interval and a preset segmentation interval to generate a plurality of candidate sets comprises:
dividing a time distribution interval according to the time interval to generate a plurality of track time intervals, wherein the time distribution interval is a distribution interval of the acquisition time of the point cloud track points;
dividing each track time interval according to the segmentation interval to generate a grouping time interval;
and grouping the point clouds according to the grouping time interval to generate the candidate sets.
4. The method according to claim 3, wherein the step of dividing the time distribution interval according to the time interval to generate a plurality of trajectory time intervals comprises:
if the time interval is larger than a preset interval threshold, determining a partitioning moment according to the acquisition moment of the point cloud track points corresponding to the time interval;
and dividing the time distribution interval according to the partitioning time to generate the track time interval.
5. The method of claim 4, wherein the interval threshold is determined based on a collection frequency of a collection vehicle.
6. The method of claim 3, wherein grouping the point clouds according to the grouping time intervals to generate the plurality of candidate sets comprises:
and distributing the point clouds corresponding to the point cloud track points of which the acquisition moments are in the same grouping time interval to the same candidate set.
7. The method according to any one of claims 1 to 6, wherein selecting an associated candidate set associated with the image from the plurality of candidate sets comprises:
acquiring a first candidate set where an associated point cloud of an image corresponding to an image acquired at the previous moment is located;
determining a second candidate set immediately following the first candidate set when ordered by acquisition time;
and taking the first candidate set and the second candidate set as association candidate sets.
8. The method of claim 7, wherein selecting the associated point cloud associated with the image from the associated candidate set comprises:
and selecting the point cloud corresponding to the point cloud track point with the minimum time as the associated point cloud associated with the image corresponding to the image track point at the current time, wherein the distance between the point cloud track point and the image track point at the current time is within a preset range in the associated candidate set.
9. A map data processing apparatus for a congested road segment, characterized by comprising:
the acquisition module is used for acquiring an image of a congested road section acquired in the daytime and a point cloud of the congested road section acquired at night;
the grouping module is used for grouping the point clouds to generate a plurality of candidate sets;
a determination module for selecting an associated candidate set associated with the image from the plurality of candidate sets;
the determination module is further configured to select an associated point cloud associated with the image from the associated candidate set;
and the generation module is used for fusing the image and the associated point cloud associated with the image to generate map data of the congested road section.
10. A processing device, comprising:
a memory for storing a program;
a processor for executing the program stored in the memory, the processor being configured to execute the map data processing method for congested road sections according to any one of claims 1 to 8 when the program is executed.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010475890.7A CN111649752B (en) | 2020-05-29 | 2020-05-29 | Map data processing method, device and equipment for congested road section |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010475890.7A CN111649752B (en) | 2020-05-29 | 2020-05-29 | Map data processing method, device and equipment for congested road section |
Publications (2)
Publication Number | Publication Date |
---|---|
CN111649752A CN111649752A (en) | 2020-09-11 |
CN111649752B true CN111649752B (en) | 2021-09-21 |
Family
ID=72349751
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202010475890.7A Active CN111649752B (en) | 2020-05-29 | 2020-05-29 | Map data processing method, device and equipment for congested road section |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111649752B (en) |
Families Citing this family (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111982018B (en) * | 2020-09-16 | 2021-11-30 | 广州天域科技有限公司 | Tracking type laser three-dimensional scanner |
CN112669250B (en) * | 2021-03-16 | 2021-09-17 | 湖北亿咖通科技有限公司 | Track alignment method and electronic equipment |
CN113362444B (en) * | 2021-05-21 | 2023-06-16 | 北京百度网讯科技有限公司 | Point cloud data generation method and device, electronic equipment and storage medium |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106845321A (en) * | 2015-12-03 | 2017-06-13 | 高德软件有限公司 | The treating method and apparatus of pavement markers information |
CN109993969A (en) * | 2019-03-08 | 2019-07-09 | 腾讯大地通途(北京)科技有限公司 | A kind of road conditions determine information acquisition method, device and equipment |
US10408939B1 (en) * | 2019-01-31 | 2019-09-10 | StradVision, Inc. | Learning method and learning device for integrating image acquired by camera and point-cloud map acquired by radar or LiDAR corresponding to image at each of convolution stages in neural network and testing method and testing device using the same |
CN110599762A (en) * | 2018-06-12 | 2019-12-20 | 光宝电子(广州)有限公司 | Road condition sensing system and method |
CN110706491A (en) * | 2019-11-05 | 2020-01-17 | 北京深测科技有限公司 | Traffic monitoring and early warning method and system for expressway |
CN110850439A (en) * | 2020-01-15 | 2020-02-28 | 奥特酷智能科技(南京)有限公司 | High-precision three-dimensional point cloud map construction method |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
AU2018278901B2 (en) * | 2018-06-14 | 2020-07-16 | Beijing Didi Infinity Technology And Development Co., Ltd. | Systems and methods for updating a high-resolution map based on binocular images |
-
2020
- 2020-05-29 CN CN202010475890.7A patent/CN111649752B/en active Active
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106845321A (en) * | 2015-12-03 | 2017-06-13 | 高德软件有限公司 | The treating method and apparatus of pavement markers information |
CN110599762A (en) * | 2018-06-12 | 2019-12-20 | 光宝电子(广州)有限公司 | Road condition sensing system and method |
US10408939B1 (en) * | 2019-01-31 | 2019-09-10 | StradVision, Inc. | Learning method and learning device for integrating image acquired by camera and point-cloud map acquired by radar or LiDAR corresponding to image at each of convolution stages in neural network and testing method and testing device using the same |
CN109993969A (en) * | 2019-03-08 | 2019-07-09 | 腾讯大地通途(北京)科技有限公司 | A kind of road conditions determine information acquisition method, device and equipment |
CN110706491A (en) * | 2019-11-05 | 2020-01-17 | 北京深测科技有限公司 | Traffic monitoring and early warning method and system for expressway |
CN110850439A (en) * | 2020-01-15 | 2020-02-28 | 奥特酷智能科技(南京)有限公司 | High-precision three-dimensional point cloud map construction method |
Also Published As
Publication number | Publication date |
---|---|
CN111649752A (en) | 2020-09-11 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111649752B (en) | Map data processing method, device and equipment for congested road section | |
US8744735B2 (en) | Traffic information generating method, traffic information generating apparatus, display, navigation system, and electronic control unit | |
US8359156B2 (en) | Map generation system and map generation method by using GPS tracks | |
CN104778274B (en) | A wide range of city road network hourage method of estimation based on sparse GPS data from taxi | |
CN109754594A (en) | A kind of road condition information acquisition method and its equipment, storage medium, terminal | |
CN108351220B (en) | Method for aggregating lane information for digital map service | |
US10740629B2 (en) | Image processing system, image processing method, information processing device and recording medium | |
CN105489008A (en) | City road congestion calculation method based on probe vehicle satellite positioning data and system | |
EP2659228B1 (en) | Generation of a seamless network of road or line segments in a digital map | |
CN102592446B (en) | Method for computing intercity road travel time by using floating car locating data | |
US20200210720A1 (en) | Non-transitory storage medium storing image transmission program, image transmission device, and image transmission method | |
CN105096590A (en) | Traffic information generation method and device | |
KR101913134B1 (en) | Apparatus for generating lane by lane traffic velocity information in intelligent trasportation system | |
US10665096B2 (en) | Non-transitory storage medium storing image transmission program, image transmission device, and image transmission method | |
CN110135216B (en) | Method and device for detecting lane number change area in electronic map and storage equipment | |
CN111337039B (en) | Map data acquisition method, device and system for congested road section and storage medium | |
CN115344655A (en) | Method and device for finding change of feature element, and storage medium | |
JP4865922B1 (en) | Driving information calculation device, navigation system, and driving information calculation method | |
CN112149471B (en) | Loop detection method and device based on semantic point cloud | |
CN114360240A (en) | High-precision positioning method based on vehicle networking track characteristics | |
EP3961156A1 (en) | Position and orientation calculation method, position and orientation calculation program, and information processing apparatus | |
CN114078322B (en) | Bus running state evaluation method, device, equipment and storage medium | |
CN115631082B (en) | Multi-scene path restoration method based on highway network topology structure | |
US11815362B2 (en) | Map data generation apparatus | |
CN112857378A (en) | Track matching method and device |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |