In-Vehicle Camera Positioning for Indoor Parking Navigation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The challenge of accurately positioning and finding a vehicle parked in an indoor parking lot due to poor GNSS signal reception, leading to difficulties in user navigation and increased complexity and cost.
Innovation Solution
A vehicle positioning method that combines vehicle travel tracking with camera-captured images to determine the vehicle's position within an indoor parking lot, utilizing semantic information to calculate and output the vehicle's location, including floor and zone information, and providing this data via SMS or applications.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If GNSS is used for vehicle positioning, then positioning accuracy is improved, but it cannot work in indoor parking lots due to poor signal reception
Solution Approach 1:
The positioning system is segmented into two independent subsystems: GNSS for outdoor positioning and visual odometry for indoor positioning. Each subsystem operates independently in its suitable environment, with the system switching between them based on location. This segmentation allows the system to maintain high positioning accuracy outdoors while gaining indoor positioning capability.
Solution Approach 2:
Visual odometry serves as an intermediary positioning method that bridges the gap between outdoor GNSS positioning and indoor positioning requirements. By using camera-captured images and semantic information as an intermediary mechanism, the system can transition from GNSS-based positioning to indoor-based positioning, enabling continuous positioning coverage across both outdoor and indoor environments.
2Adaptability or versatility
If visual odometry with semantic information is used for indoor positioning, then indoor positioning capability is improved, but positioning cost increases
Solution Approach 1:
The camera system performs multiple functions: it captures images for visual odometry positioning, extracts semantic information for positioning, and can potentially be used for other vehicle functions. By making the camera multi-functional, the system avoids adding dedicated indoor positioning hardware, thereby reducing overall system cost while maintaining indoor positioning capability.
Solution Approach 2:
The system uses the vehicle's existing camera and processor to perform visual odometry and semantic information extraction for positioning. Instead of requiring external positioning infrastructure or dedicated positioning hardware, the vehicle serves its own positioning needs using its onboard resources, reducing dependency on external systems and lowering overall positioning system cost.
3Speed
If visual odometry is used for indoor positioning, then positioning speed is improved, but processing complexity increases
Solution Approach 1:
The system pre-extracts semantic information from captured images and pre-calculates feature points for visual odometry. By performing these processing tasks in advance rather than in real-time during positioning, the system reduces real-time processing complexity while maintaining fast positioning speed. The pre-processed data can be quickly referenced during actual positioning operations.
Data Source
AI summary
A vehicle positioning method is provided, includes: when a vehicle travels into an indoor parking lot, capturing, by using a camera in the vehicle, an image at a track point included on a traveling track of the vehicle in a traveling process of the vehicle; obtaining a first traveling track of the vehicle on a target floor in the indoor parking lot, where the target floor is a floor on which the vehicle is currently located; calculating position information of a current position of the vehicle in the indoor parking lot based on semantic information included in an image at a track point on the first traveling track, where the position information includes a floor and/or a zone; and when the current position of the vehicle is a parking space, outputting position information of the parking space in the indoor parking lot.