Autonomous Vehicle Road Object Detection Using Camera-LiDAR Fusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in detecting unknown objects on road surfaces due to limitations in LiDAR methods with low resolution scan points and high false positives in image data analysis, particularly with machine learning models trained on small datasets.
Innovation Solution
A computer system integrates image processing techniques, including panoptic segmentation and machine learning models, to identify potential unknown objects, and verifies these detections using LiDAR scans to generate accurate 3D bounding boxes for unknown objects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If LiDAR methods are used for object detection, then detection coverage is improved, but measurement precision deteriorates due to low resolution scan points
Solution Approach 1:
The patent combines LiDAR and camera data through sensor fusion, where LiDAR provides broad coverage and the camera provides high-resolution details. The system integrates point cloud data from LiDAR with image data from the camera to achieve both wide detection coverage and high measurement precision for unknown objects on the road surface.
2Measurement precision
If image data is used for object detection, then measurement precision is improved, but reliability deteriorates due to false positives
Solution Approach 1:
The system uses LiDAR data as feedback to verify camera-based object detections. The sensor fusion module cross-references the high-resolution image data with the spatial information from LiDAR to confirm whether detected objects are genuine road surface objects or false positives, thereby improving reliability while maintaining measurement precision.
3Device complexity
If machine learning models are trained on small datasets, then device complexity is reduced, but measurement precision deteriorates
Solution Approach 1:
The patent introduces sensor fusion as an intermediary that combines data from multiple sources (LiDAR and camera) to compensate for the limitations of small training datasets. This intermediary mechanism allows the system to achieve high detection accuracy without requiring extensive training data, as the complementary information from both sensors enriches the input for the machine learning model.
4Ease of operation
If 2D image data is used for object detection, then ease of operation is improved, but measurement precision deteriorates due to inability to distinguish depth
Solution Approach 1:
The patent transitions from 2D image data to 3D spatial understanding by integrating LiDAR point cloud data. This adds the depth dimension to the detection system, enabling accurate determination of object distance, elevation, and spatial relationships while maintaining the computational efficiency of image-based processing through coordinated sensor fusion.
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 enhances the ability of autonomous vehicles to efficiently detect and navigate around unknown objects, improving safety and route planning by reducing false positives and ensuring accurate object classification.
Implementation Method 1
retrieving, by the processor, a set of data points received from a LiDAR sensor of the autonomous vehicle monitoring the environment of the autonomous vehicle
Data Source
AI summary
An autonomous vehicle comprises one or more processors. The processors can be configured to receive, from a sensor of the autonomous vehicle, an image of an environment outside of the autonomous vehicle. The processors can detect potential unknown objects based on the image. The processors can compare the detection based on the image to a set of data points of a LiDAR scan to determine if there are unknown objects on a roadway.