Driving Assistance Ground Modeling for Real-Time Obstacle Separation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies for perceptive autonomous vehicles struggle to accurately distinguish between ground and obstacles in real-time using data point clouds from LIDAR or stereo cameras, with previous methods being either inaccurate or computationally costly.
Innovation Solution
A computerized device and method employing a Gaussian conditional random field with spatial and temporal components, utilizing an expectation-maximization algorithm to calculate the probability of data points belonging to a reference surface, enabling real-time differentiation between ground and obstacles with calculational costs compatible with autonomous vehicle processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If Markov conditional fields or MRF are used to determine ground and obstacles, then measurement precision is improved, but productivity deteriorates due to non real-time processing
Solution Approach 1:
The patent changes the mathematical model parameters from Markov conditional fields to Gaussian conditional random fields, which allows for real-time computation while maintaining high accuracy in ground-obstacle distinction through probabilistic modeling of point cloud data
Solution Approach 2:
The patent replaces the computationally intensive MRF system with a Gaussian conditional random field system that uses statistical properties and probability distributions to achieve the same ground-obstacle separation function with reduced computational overhead
2Productivity
If plane fitting technique is used to separate ground and obstacles, then productivity is improved for real-time use, but measurement precision deteriorates
Solution Approach 1:
The patent transitions from deterministic plane fitting parameters to probabilistic Gaussian distribution parameters, allowing the system to model the statistical variability of ground points and obstacles, thereby improving measurement precision while maintaining real-time processing through efficient probability calculations
3Measurement precision
If 3D LIDAR or stereo cameras are used to increase data points, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The patent introduces a Gaussian conditional random field as an intermediary computational layer that processes the raw point cloud data from 3D LIDAR or stereo cameras, transforming the complex sensor data into structured ground-obstacle classifications through probabilistic modeling, thereby managing device complexity while preserving measurement precision
Data Source
AI summary
A computerized device for driving assistance comprises a memory (4) designed to receive data point cloud data (8) in which a point cloud associates, for a given instant, points each having coordinates in a plane associated with the point cloud and a value denoting a height. The device furthermore comprises a calculator (6) designed to access the memory (4) and, for a given point cloud, to calculate data on the probability of belonging to a reference surface, associated with each point of the data point cloud, on the one hand, and node data associating a value denoting a height (hi) and two values indicating a slope in a plane associated with the plane of the given point cloud, on the other hand, by determining a Gaussian random conditional field by way of the data point cloud data (8) corresponding to the given point cloud, which Gaussian random conditional field is represented by a mesh of nodes in said associated plane, which nodes are defined by the node data, and to return the data on the probability of belonging to a reference surface and/or at least some of the node data and values denoting a height.