Multi-Camera Frame Rate Allocation for Vehicle Perception
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing automatic driving systems face challenges in setting optimal frame rates for multiple cameras, leading to increased data processing loads and reduced accuracy in environmental perception, especially when controlling vehicles in a convoy or group.
Innovation Solution
Implementing a system with a first processor outputting point information from a high-frame-rate camera and a second processor outputting identification information from a lower-frame-rate camera, with a third processor associating this data to manage processing loads and enhance environmental awareness.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple cameras image the surroundings of the vehicle at high frame rates, then the accuracy of environmental perception is improved, but the amount of data and calculation required for control increases
Solution Approach 1:
The patent segments the camera system into multiple cameras with different frame rates based on their functional requirements. The first camera captures images at a first frame rate while the second camera captures at a second frame rate, allowing the system to distribute the imaging load and reduce overall data quantity while maintaining perception accuracy for critical areas
Solution Approach 2:
Different cameras are assigned different frame rates according to their specific imaging requirements and importance. The system applies local quality by making the first camera operate at a higher frame rate for critical regions while the second camera operates at a lower frame rate for less critical regions, optimizing the balance between perception accuracy and data management
2Measurement precision
If multiple cameras image the surroundings of the vehicle at high frame rates, then the accuracy of environmental perception is improved, but the calculation load for control increases
Solution Approach 1:
The patent segments the camera system into multiple cameras with different frame rates based on their functional requirements. The first camera captures images at a first frame rate while the second camera captures at a second frame rate, allowing the system to distribute the imaging load and reduce overall data quantity while maintaining perception accuracy for critical areas
Solution Approach 2:
Different cameras are assigned different frame rates according to their specific imaging requirements and importance. The system applies local quality by making the first camera operate at a higher frame rate for critical regions while the second camera operates at a lower frame rate for less critical regions, optimizing the balance between perception accuracy and data management
3Reliability
If a lateral camera is mounted on each vehicle in a convoy to recognize lateral conditions, then the safety of convoy traveling is improved, but the processing load on the information processing apparatus increases significantly
Solution Approach 1:
The patent merges the imaging function for lateral conditions into a shared resource model where a single lateral camera serves multiple vehicles in the convoy. The information processing apparatus processes images from this shared lateral camera to recognize conditions on lateral sides for the entire convoy, eliminating the need for each vehicle to have its own lateral camera and thereby significantly reducing the processing load while maintaining safety
Solution Approach 2:
The lateral camera is designed as a universal imaging device that can serve multiple vehicles in the convoy simultaneously. This multi-functional approach allows one camera to perform the safety monitoring function for the entire convoy rather than requiring dedicated cameras for each vehicle, optimizing resource utilization and reducing processing demands
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An information processing apparatus includes: a first processor that outputs, from an image of an object captured by a first camera, point information in which the imaged object is captured as a point; a second processor that outputs, from an image of the object captured by a second camera directed in a direction corresponding to the first camera, identification information that identifies the imaged object; and a third processor that associates the point information with the identification information, and a frame rate of the first camera is higher than a frame rate of the second camera.