Pavement Element Annotation via Point Cloud to Overhead Image Fusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional methods for annotating pavement elements in point cloud data are labor-intensive, require high hardware resources, and often result in incomplete annotations due to the complexity of processing point cloud data and the need to consider X, Y, and Z dimensions simultaneously.
Innovation Solution
A pavement element annotation method that constructs single-frame point clouds into a joint point cloud in a global coordinate system, removes dynamic objects, transforms the data into an overhead image, pre-annotates pavement elements using a pre-annotation model, and establishes a transformational relation between the overhead image and the point cloud data using a ball query algorithm and a ground point algorithm.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional manual annotation methods are used for pavement elements, then annotation accuracy can be maintained, but annotation time and labor intensity increase significantly
Solution Approach 1:
The patent applies preliminary action by performing automatic pre-annotation using a pre-trained model before manual annotation. The model generates initial annotation results that annotators then refine, significantly reducing the time and labor required for complete annotation while maintaining accuracy through the combination of automated preprocessing and human review.
2Measurement precision
If high-precision point cloud data from lidar is used for annotation, then annotation accuracy improves, but hardware requirements and processing complexity increase
Solution Approach 1:
The patent extracts and removes dynamic objects from the point cloud data before annotation, isolating only the static pavement elements that need to be annotated. This reduction in data complexity decreases hardware requirements and processing complexity while maintaining annotation accuracy by focusing computational resources on the relevant static elements.
3Ease of operation
If single-frame point cloud data is annotated directly, then annotation process is simpler, but annotation completeness decreases due to occlusion and holes
Solution Approach 1:
The patent merges multiple consecutive frames of point cloud data into a unified annotation result. By combining information from multiple frames, the system overcomes occlusion and holes present in individual frames, achieving complete annotation of pavement elements while maintaining operational simplicity through automated multi-frame processing.
4Measurement precision
If annotators consider X, Y, and Z dimensions simultaneously in point cloud annotation, then annotation accuracy is maintained, but annotation efficiency decreases
Solution Approach 1:
The patent segments the three-dimensional annotation task into two distinct stages: first, the pre-trained model performs automatic annotation in 3D space, and second, annotators perform manual refinement primarily in the two-dimensional overhead view. This segmentation reduces the dimensional complexity annotators must handle simultaneously, improving efficiency while maintaining accuracy through the combined approach.
Data Source
AI summary
The present invention discloses a pavement element annotation method for point cloud data with fusion of height, which comprises the following steps: constructing all single-frame point clouds into a joint point cloud in a global coordinate system based on the pose of each frame of single-frame point cloud; removing dynamic objects in the joint point cloud to obtain a static joint point cloud; transforming the static joint point cloud into an overhead image; pre-annotating the pavement elements in an overhead image by using the pavement element pre-annotation model; modifying the pre-annotated result to obtain an overhead image annotation; based on a ball query algorithm and a ground point algorithm, establishing a transformational relation between the pixels of the overhead image and the points of the static joint point cloud; transforming the overhead image annotation into a static joint point cloud annotation based on the transformational relation.


