Recognition method, device, electronic device and storage medium of forbidden road

Abstract

Embodiments of the present application provide a method, device, electronic device, and storage medium for identifying forbidden roads, which relate to the technical field of data mining and can be used for identifying forbidden roads in maps. The method includes: acquiring road network data, constructing a topological map according to the road network data, determining identified prohibited roads from the road network data, traversing the topological map starting from the line segment corresponding to the identified prohibited roads, and determining the currently traversed According to the topology structure corresponding to the line segment, and obtain the track data of at least one adjacent line segment; according to the topology structure and the track data of at least one adjacent line segment, search for the adjacent line segment belonging to the prohibited road in at least one adjacent The adjacent segment of the road determines the segment to be traversed next. The embodiment of the present application has a higher accuracy rate than the existing method of identifying prohibited roads based on single trajectory data and image recognition.

Description

technical field [0001] The present application relates to the technical field of data mining, and in particular, the present application relates to a method for identifying forbidden roads. Background technique [0002] Forbidden roads for pedestrians and bicycles refer to road sections where the passage of pedestrians, bicycles and electric vehicles is restricted within a specified time for reasons such as maintaining traffic order, passing safety, and city appearance. Pedestrian and cycling prohibition is the traffic capacity attribute of the road. It is expressed in the data as an additional attribute of the road, rather than a line-point-line relationship. It needs to be distinguished from the traffic limit. [0003] There are many factors that affect walking and cycling navigation, among which the problem with a relatively high proportion is the lack of prohibited roads for walking and cycling. Danger. It is difficult to identify the forbidden roads for walking and ri...

AI Technical Summary

Problems solved by technology

[0013] 1) The quality of the track is poor. The data of the walking track is different from the data of the car track. Often, the users of the walking track do not necessarily abide by the traffic rules, and there may be retrograde or disorderly traffic, which results in the big data of the original walking track. contains a lot of noise data, which may include: not clearly distinguishing the two sources of data, including walking, cycling, etc.; track point drift and loss due to equipment reasons; track point abnormality and loss due to environmental reasons; user Abnormal track points caused by abnormal behavior, such as parking, detour, retrograde, and illegal driving

Due to the above special reasons, it is necessary to perform a strong degree of trajectory preprocessing on the trajectory, but there may still be relatively abnormal trajectory data

[0014] 2) The trajectories cannot be fully covered. Based on the application of the existing map business, it is impossible to completely match the trajectories of users in the real world. The focus is on some relatively remote paths or village roads and parallel roads. There are certain limitations

[0017] 1) There are few identifiable elements, and the identifiable forbidden scene is single;

[0018] 2) The number of identifiable roads is small, and the road grades are small. The roads recognized by images are often prohibited roads collected from high-level road signs in some cities. Some prohibited roads cannot be recognized without making prohibited signs ; At the same time, due to the low coverage of the collected crowdsourcing images, there are few identifiable roads;

[0019] 3) The cycle of the image recognition technology solution is long, and the state of the identified road is likely to change. From the collection of signboards to the identification of prohibited roads for walking and riding, and finally the verification of the identified roads, the cycle is relatively long

The collection cost is high, the cycle is long, the amount of trajectory mining data is small, the image recognition cycle and recognition amount are small, and the difficulty of verification after recognition has caused serious problems in the early stage of the project.

Images