LiDAR Point Cloud Filling for Missing Pixels and Edge Preservation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

LiDAR systems suffer from low accuracy in point cloud filling due to environmental noise and noise interference, leading to missing points in the point cloud data, which affects recognition accuracy.

Innovation Solution

A point cloud filling method that determines the position of missing pixels in LiDAR data, calculates a point filling mark based on smoothing and concentration features, and fills the missing pixels with accurate distance and reflectivity values to maintain object edge features and improve recognition accuracy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If point cloud filling is performed on all missing points, then the completeness of point cloud data is improved, but the accuracy of filling decreases due to noise interference and environmental factors

Engineering Contradiction:
Improvecompleteness of point cloud dataVSAvoidaccuracy of point cloud filling
Core Design Contradiction:
ReliabilityVSMeasurement precision

Solution Approach 1:

The patent applies local quality by differentiating between different types of missing points and applying different filling strategies. Points on object surfaces are filled using surface interpolation, while points in empty spaces are left unfilled. This selective approach ensures that filling operations are performed only where they improve completeness without compromising accuracy.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The patent performs preliminary classification of missing points before filling. By first identifying whether a missing point lies on an object surface or in empty space, the system prepares appropriate filling strategies in advance, preventing inaccurate filling of points that should remain empty.

Inventive Principle:
Principle #10Preliminary action

2Reliability

If traditional 3D point cloud filling methods are used, then the completeness of point cloud data is improved, but the complexity of the filling process increases and edge features may be deformed

Engineering Contradiction:
Improvecompleteness of point cloud dataVSAvoidcomplexity of point cloud filling process
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent segments the point cloud filling process into distinct stages: identifying missing points, classifying their locations relative to object surfaces, and applying appropriate filling methods. This segmentation simplifies the overall complexity by breaking down the filling process into manageable, independent steps.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Instead of attempting to fill all missing points uniformly, the patent inverts the approach by selectively identifying and filling only those missing points that lie on object surfaces. This inversion of the traditional approach reduces complexity by avoiding unnecessary filling operations in empty spaces while preserving edge features.

Inventive Principle:
Principle #13The other way round (Inversion)

3Productivity

If point cloud filling is performed without selective judgment, then the processing speed is improved, but the recognition accuracy decreases due to contour deformation

Engineering Contradiction:
Improveprocessing speed of point cloud fillingVSAvoidrecognition accuracy of LiDAR
Core Design Contradiction:
ProductivityVSMeasurement precision

Solution Approach 1:

The patent performs preliminary classification of missing points to determine their location relative to object surfaces before executing filling operations. This preliminary action enables the system to quickly identify which points require filling, maintaining processing speed while ensuring that only appropriate points are filled, thus preserving recognition accuracy.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent applies local quality by using different filling approaches for different locations. Points on object surfaces are filled using surface interpolation to maintain edge features, while points in empty spaces are left unfilled. This localized approach ensures high recognition accuracy without sacrificing processing speed.

Inventive Principle:
Principle #3Local quality

Data Source

PatentEP4517655B1Point cloud filling method, device, equipment, and storage medium
Publication Date: 2026.03.11 SUTENG INNOVATION TECHNOLOGY CO LTD
  • EP4517655B1 patent drawingFigure 1~2
  • EP4517655B1 patent drawingFigure 3~4
  • EP4517655B1 patent drawingFigure 5

AI summary

The present application provides a point cloud filling method, device, equipment, and storage medium. The method includes: obtaining the point cloud data collected by the LiDAR and determining the point missing pixel position from the point cloud data; obtaining the point filling mark of the point missing pixel position; determining the point missing pixel position as the point filling pixel position; performing point cloud filling on the point filling pixel position according to the point filling mark. In this way, can effectively improve the accuracy of point cloud filling, maintain the edge features of the object being measured in the field of view, and not cause contour deformation of the object being measured due to point cloud filling, thereby improving the accuracy of LiDAR recognition.