Point Cloud Weighting and Ray Casting for Clutter Removal
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
During three-dimensional scanning, invalid unwanted objects are often scanned along with the target objects, forming noise data that can interfere with subsequent processing.
Innovation Solution
A point cloud processing method that utilizes a global model to assign weights to three-dimensional points, projects the current frame into this model, casts light rays to identify noise points, reduces their weights, and deletes them if they meet a noise-point deletion condition, ensuring accurate deletion through multiple verifications.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If three-dimensional scanning is performed to capture spatial coordinates of objects, then digitalization of physical objects is achieved, but noise data from unwanted objects is also captured and interferes with subsequent processing
Solution Approach 1:
The system performs preliminary actions by maintaining a global model that accumulates point cloud data from historical frames before processing the current frame. This pre-established context allows for proactive noise identification and removal, preventing noise from interfering with subsequent processing steps.
Solution Approach 2:
The system uses feedback mechanisms by comparing the current frame's point cloud against the accumulated global model. The weight adjustment process provides continuous feedback, where points consistent with the global model receive higher weights and points deviating from it receive lower weights, enabling iterative refinement of noise filtering.
2Manufacturing precision
If noise points are deleted from the point cloud, then processing quality is improved, but risk of erroneous deletion of valid points increases
Solution Approach 1:
The system changes parameters by assigning and adjusting weights for each point based on multiple criteria including depth information, normal vector consistency, and density distribution. This multi-parameter approach allows for nuanced differentiation between noise points and valid points, reducing erroneous deletions while maintaining high processing accuracy.
Solution Approach 2:
The global model serves as a preliminary reference that accumulates valid point cloud data before the current frame is processed. This pre-established model provides a baseline for comparison, enabling the system to identify noise points more reliably without mistakenly deleting valid points that should be preserved.
3Measurement precision
If multiple verifications are performed to reduce weights of target points, then deletion accuracy is enhanced, but processing time increases
Solution Approach 1:
The system applies partial action by selectively processing only those points that require verification against the global model. Not all points undergo the full multi-stage weight reduction process; only points identified as potential noise candidates based on initial filtering criteria are subjected to extensive verification, optimizing the balance between accuracy and processing time.
4Reliability
If global model is constructed by fusing point clouds of historical frames, then noise filtering capability is improved, but computational complexity increases
Solution Approach 1:
The system extracts only the essential and relevant features from historical point cloud frames to construct the global model, rather than storing and processing complete point cloud data. This extraction approach maintains noise identification accuracy while significantly reducing computational complexity and memory requirements.
Data Source
AI summary
The present application provides a point cloud processing method and apparatus, a device, and a storage medium. The method comprises: acquiring a current frame point cloud collected by a point cloud collection apparatus and a global model, wherein the global model is obtained by fusing historical frame point clouds collected by the point cloud collection apparatus, any three-dimensional point in the global model carries a weight, and the weight of any three-dimensional point represents the possibility that the three-dimensional point is a cluttered point or a target object; after the current frame point cloud is projected into the global model, projecting a plurality of light rays between the projected current frame point cloud and the point cloud collection apparatus to determine a target three-dimensional point through which the light rays pass in the global model; reducing the weight of the target three-dimensional point; and deleting the target three-dimensional point if the reduced weight of the target three-dimensional point satisfies a cluttered point deletion condition. Therefore, the cluttered point is accurately deleted.