Integrated Vision-LiDAR Sensor Layout for Consistent Vehicle Perception
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Intelligent driving vehicles face challenges due to environmental susceptibility of vision cameras, high cost of LiDAR, limited data acquisition with single sensors, and inconsistent fields of view among scattered sensor layouts.
Innovation Solution
An integrated sensor comprising a binocular vision camera and LiDAR with aligned pixel-level scanning, a long-range vision camera, and a control circuit board for comprehensive data analysis, mounted in a housing with sealing and heat dissipation features, and a mobile platform design.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If multiple sensors are scattered mounted on the vehicle (roof, windshields, mirrors), then the coverage area is expanded, but the fields of view become inconsistent and data analysis becomes difficult
Solution Approach 1:
The patent combines multiple sensors (LiDAR, binocular cameras, monocular cameras, infrared cameras) into a single integrated sensor assembly mounted at one location on the vehicle. This merging approach maintains comprehensive coverage while ensuring consistent fields of view across all sensors, as they all observe from the same position and orientation.
2Measurement precision
If LiDAR is used for distance measurement, then measurement precision is improved, but the cost increases significantly
Solution Approach 1:
The patent integrates LiDAR with vision cameras in a unified sensor system. By combining LiDAR's precise distance measurement with cameras' texture and color perception capabilities, the system achieves comprehensive environmental understanding. The integrated design allows cost optimization through shared mounting structures, power supply, and data processing resources while maintaining the precision benefits of LiDAR.
3Ease of manufacture
If vision cameras are used for environmental perception, then the cost is reduced, but environmental susceptibility increases
Solution Approach 1:
The patent merges vision cameras with complementary sensors (LiDAR for distance, infrared cameras for thermal information) to create a robust multi-sensor system. This combination allows the system to overcome the environmental limitations of individual camera types by using multiple sensing modalities that are less susceptible to specific environmental conditions, thereby improving overall reliability while maintaining cost-effectiveness.
4Device complexity
If a single sensor type is used, then device complexity is reduced, but data acquisition capability becomes limited
Solution Approach 1:
The patent creates a universal sensor assembly that performs multiple functions: LiDAR provides distance measurement, binocular cameras provide depth perception and visual recognition, monocular cameras provide wide-area surveillance, and infrared cameras provide thermal information. This multi-functional integrated system comprehensively captures environmental information without requiring separate sensor systems for each function.
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
Enhances perception accuracy and reduces costs by aligning sensor data, providing comprehensive environmental data, and overcoming sensor layout inconsistencies.
Implementation Method 1
a ranging module of the LiDAR is disposed inside the cavity and is configured to emit and receive a laser beam which exits and returns through the optical window
Data Source
Figure 1~2
Figure 3~4
Figure 5~6
AI summary
An integrated sensor including a housing with an internal cavity; a control circuit board disposed inside the cavity; a binocular vision camera disposed on the housing and including a left-eye camera and a right-eye camera electrically connected to the control circuit board; and a LiDAR electrically connected to the control circuit board, wherein an optical window of the LiDAR is disposed on the housing, and a ranging module of the LiDAR is disposed inside the cavity and is configured to emit and receive a laser beam which exits and returns through the optical window. The integrated sensor provided herein can be mounted externally or internally on a vehicle, overcoming drawbacks of scattered sensor layouts and inconsistent fields of view. The binocular vision camera provides the vehicle with visual image information of the surrounding environment, the LiDAR provides the vehicle with three-dimensional point cloud data, and the control circuit board performs comprehensive analysis of the visual image information and three-dimensional point cloud data, thereby meeting the growing demand of intelligent driving.