Point Cloud Group Synchronization for Sensor Misalignment

Solution Overview

Problem

Existing methods for connecting point cloud data with related data, such as those using LIDAR and image sensors, often result in increased errors due to asynchronous data acquisition times, leading to misalignment of object positions and inaccurate distance calculations.

Innovation Solution

A method that involves classifying point cloud data into groups, assigning labels, predicting moving routes, and synchronizing the groups with the acquisition time of related data, thereby reducing positional misalignment and error in distance calculations by replacing groups with positions at the same time as the related data.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If point cloud data and related data are acquired asynchronously using LIDAR and image sensors, then data acquisition efficiency is improved, but positional alignment accuracy deteriorates

Engineering Contradiction:

Improvedata acquisition efficiencyVSAvoidpositional alignment accuracy

Core Design Contradiction:

ProductivityVSMeasurement precision

Solution Approach 1:

The system performs preliminary classification of point cloud data into groups and predicts moving routes before synchronization. By pre-processing the point cloud data to identify objects and their trajectories, the system can later accurately map these pre-classified groups to image data even when acquisition times differ, thus maintaining positional accuracy while preserving asynchronous acquisition efficiency

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent introduces an intermediate processing layer that includes group classification, label assignment, and moving route prediction. This intermediary process acts as a bridge between asynchronously acquired point cloud data and image data, enabling accurate temporal and spatial synchronization without requiring simultaneous data capture, thereby resolving the contradiction between acquisition efficiency and alignment accuracy

Inventive Principle:

Principle #24Intermediary (Mediator)

2Speed

If point cloud data from different acquisition times are directly connected with related data, then processing speed is improved, but distance calculation accuracy deteriorates

Engineering Contradiction:

Improveprocessing speedVSAvoiddistance calculation accuracy

Core Design Contradiction:

SpeedVSMeasurement precision

Solution Approach 1:

The system performs preliminary classification of point cloud data into groups and predicts moving routes before synchronization. By pre-processing the point cloud data to identify objects and their trajectories, the system can later accurately map these pre-classified groups to image data even when acquisition times differ, thus maintaining positional accuracy while preserving asynchronous acquisition efficiency

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent dynamically adjusts the synchronization process by predicting moving routes of objects and using these predictions to accurately map point cloud groups to image data at different acquisition times. This dynamic approach allows the system to handle time-varying positional information correctly, maintaining distance calculation accuracy while enabling efficient processing of asynchronous data without requiring direct one-to-one temporal matching

Inventive Principle:

Principle #15Dynamics

Data Source

PatentUS20240095932A1Method, non-transitory computer readable storage medium, training data set, and device for connecting point cloud data with related data

Publication Date: 2024.03.21 DENSO CORP

AI summary

A point cloud data is connected with a related data. Multiple sets of point cloud data are prepared. Each point cloud data includes information of a point cloud connected to three-dimensional position information, and each point cloud data is connected to acquisition time. At least one group is generated by classifying the point cloud, and the group is assigned to a position label and a moving body label. A moving route of the group with a moving body on-flag is predicted based on the position label of the group. The group is replaced with a position at acquisition time of the related data according to the moving route, and the acquisition time of the related data is connected to the group.