Point cloud map construction method and device, equipment and storage medium
A map construction and point cloud technology, applied in the field of point cloud maps, can solve the problems of tailing and poor map rendering effect, and achieve the effect of avoiding tailing
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0031] figure 1 It is a flow chart of a point cloud map construction method provided by Embodiment 1 of the present invention. This embodiment is applicable to mapping scenarios such as city simulation and high-precision map construction. The purpose is to realize mapping of static scenes based on point clouds. At the same time, the trailing phenomenon caused by the temporarily existing dynamic objects in the static scene is eliminated, and the dynamic objects refer to objects that keep moving. The static scene specifically refers to a scene that does not include movable objects, and the movable objects refer to objects that have motion attributes but do not necessarily move, such as pedestrians and vehicles. The specific movable object is an object customized according to the application scenario. For example, in the high-precision map production industry, pedestrians, buses, cars, and bicycles in the street view scene can all be defined as movable objects; in the indoor scen...
Embodiment 2
[0058] image 3 It is a flowchart of a point cloud map construction method provided by Embodiment 2 of the present invention. On the basis of the above embodiments, this embodiment provides the segmentation results of the same instance under multiple continuous single-frame point clouds, and the same The process of determining the state of the instance in each single-frame point cloud realizes the accurate determination of the state of the instance, and provides a reliable basis for determining the target patching algorithm. The explanations of terms that are the same as or corresponding to those in the foregoing embodiments are not repeated here.
[0059] see image 3 The point cloud map construction method provided in this embodiment specifically includes the following steps:
[0060] Step 310 , for multiple consecutive single-frame point clouds of the scene to be mapped, perform instance segmentation based on the current single-frame point cloud, and obtain an instance se...
Embodiment 3
[0118] Figure 4 It is a flowchart of a method for constructing a point cloud map provided by Embodiment 3 of the present invention. This embodiment is further optimized on the basis of the foregoing embodiments. Specifically: if hole patching is only performed in the world coordinate system, whether using the multi-view projection patching algorithm or the deep learning patching algorithm for patching, the patching results are often dense and uneven, because the multi-view projection patching algorithm will combine all neighbors All the relevant point cloud points in a single frame point cloud are projected into the convex bounding box, and the deep learning patching algorithm has a probabilistic deviation, so the point cloud points in the patched hole area are dense and uneven, and the visual effect is poor. In view of this problem, this embodiment further provides a solution for performing point cloud thinning on the patched hole area. The explanations of terms that are th...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More 


