Coherent LiDAR Point Cloud Segmentation Using Doppler Velocity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing autonomous vehicle systems face challenges in accurately and efficiently segmenting point clouds to identify and track moving objects, particularly due to the limitations of standard Time-of-Flight (ToF) lidar technology in determining velocities based on a single sensing frame.
Innovation Solution
The implementation of Doppler-assisted segmentation using coherent lidar technology, which leverages the Doppler effect to associate radial velocity with return points in the point cloud, enabling efficient segmentation of objects with distinct velocity signatures even in a single sensing frame.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If standard Time-of-Flight (ToF) lidar technology is used, then the system can obtain point cloud data, but it cannot accurately determine velocities of objects based on a single sensing frame
Solution Approach 1:
The patent transitions from standard ToF lidar to coherent lidar technology, changing the measurement parameters from only time-of-flight to include both time-of-flight and Doppler frequency shift. This parameter change enables velocity determination from a single sensing frame by extracting radial velocity information through Doppler effect measurement, directly resolving the contradiction between measurement precision and information loss.
2Loss of information
If coherent lidar technology with Doppler effect is implemented, then velocity information can be obtained from single sensing frame, but device complexity increases
Solution Approach 1:
The coherent lidar system performs multiple functions simultaneously: it measures both distance (through time-of-flight) and radial velocity (through Doppler frequency shift) using the same hardware platform. This multi-functionality reduces the need for separate sensing systems and mitigates the complexity increase by consolidating multiple measurement capabilities into a single device.
3Reliability
If point cloud segmentation is performed without velocity information, then processing is simpler, but objects with different velocities cannot be differentiated when closely positioned
Solution Approach 1:
The patent applies segmentation to the point cloud data based on velocity characteristics, dividing the point cloud into distinct groups corresponding to objects with different radial velocities. This velocity-based segmentation enables reliable differentiation of closely positioned objects with different velocities, directly improving object identification accuracy while using the velocity information already obtained through coherent lidar.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach allows for accurate identification and tracking of moving objects by effectively differentiating between objects with different velocities, even when closely positioned, and correctly identifying extended parts of point clouds as belonging to a single object.
Implementation Method 1
The implementation of Doppler-assisted segmentation using coherent lidar technology, which leverages the Doppler effect to associate radial velocity with return points in the point cloud
Data Source
AI summary
Aspects and implementations of the present disclosure address shortcomings of the existing technology by enabling Doppler-assisted segmentation of points in a point cloud for efficient object identification and tracking in autonomous vehicle (AV) applications, by: obtaining, by a sensing system of the AV, a plurality of return points comprising one or more velocity values and one or more coordinates of a reflecting region that reflects a signal emitted by the sensing system, the one or more velocity values and the one or more coordinates obtained for the same instance of time, identifying that the set of the return points is associated with an object in an environment, and causing a driving path of the AV to be determined in view of the object.