Autonomous Vehicle Localization Using Cameras Instead of LiDAR
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing autonomous driving systems for short-distance commuter vehicles rely on mechanical LiDARs, which are costly, complex, and prone to damage, making them unsuitable for small, lightweight vehicles like kick scooters and self-balancing scooters.
Innovation Solution
An autonomous vehicle driving system that uses a sensing device with an image sensor to capture environmental images, a locating device with a vehicle-side locator based on differential operations to determine the vehicle's location, and a control device to manage vehicle driving based on the determined location and environmental data, eliminating the need for LiDAR.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If mechanical LiDAR is used for autonomous driving, then autonomous driving capability is achieved, but cost increases and device complexity increases
Solution Approach 1:
The patent replaces the mechanical LiDAR system with a combination of image sensor (camera) and localization device. The image sensor captures environmental images while the localization device determines vehicle position through differential operations, eliminating the need for complex mechanical LiDAR hardware while maintaining autonomous driving functionality.
Solution Approach 2:
Instead of using direct mechanical LiDAR for spatial perception, the system creates a digital representation of the environment through image capture and processes localization data to reconstruct positional information. This copying approach using visual and computational methods achieves similar functional outcomes without the mechanical complexity.
2Reliability
If mechanical LiDAR is used for autonomous driving, then autonomous driving capability is achieved, but vehicle cost increases
Solution Approach 1:
The patent employs image sensors and standard localization components that are significantly cheaper than mechanical LiDAR systems. These components are mass-producible and can be easily integrated into short-distance commuter vehicles, making autonomous driving economically viable for budget-friendly vehicle segments.
Solution Approach 2:
By substituting expensive mechanical LiDAR with more affordable image-based perception and computational localization systems, the patent dramatically reduces the bill of materials cost while maintaining the essential autonomous driving capability.
3Reliability
If mechanical LiDAR is used for autonomous driving, then autonomous driving capability is achieved, but vehicle size increases
Solution Approach 1:
The patent replaces the bulky mechanical LiDAR assembly with compact image sensors and electronic localization devices. This substitution dramatically reduces the physical space required for sensing components, making autonomous driving systems suitable for compact short-distance commuter vehicles where space is at a premium.
Data Source
AI summary
This disclosure provides an autonomous vehicle driving system and method. A vehicle may be located by using a vehicle-determined location determined by a vehicle-side locator based on a differential operation and an environmental image captured by an image sensor, that is, a current vehicle location of the vehicle is determined, so that accuracy of vehicle locating is ensured. Further, vehicle driving may be controlled based on the current vehicle location and environmental sensing data, and thus autonomous driving can be implemented without a LiDAR during vehicle operation, thereby reducing costs without affecting the appearance and use of the vehicle.