Ground-Plane Object Positioning With Successive Road-Height Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing object detection and tracking systems in autonomous vehicles face inaccuracies due to imprecise road-height models, leading to incorrect determination of object positions, which can impact navigation and trajectory planning.
Innovation Solution
The system employs a method to determine object positions by projecting vectors from a camera center point through an image representation of an object onto a road-height model, updating these positions using successive road-height models obtained from LIDAR or RADAR captures to enhance accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If object detection systems use traditional road-height models, then the system complexity remains low, but the measurement precision of object positions deteriorates
Solution Approach 1:
The system performs preliminary actions by capturing multiple images at different camera positions before final object position determination. Road-height models are updated in advance using these preliminary captures, allowing the final position calculation to benefit from more accurate, pre-computed geometric information without increasing real-time processing complexity
Solution Approach 2:
The system implements feedback by iteratively updating road-height models based on captured images and comparing determined object positions against the updated models. This feedback loop continuously refines the road-height model accuracy, which in turn improves subsequent object position determinations
2Reliability
If the system uses a single road-height model, then the processing time is reduced, but the reliability of object position determination deteriorates
Solution Approach 1:
Multiple road-height models are generated in advance using preliminary image captures at different camera positions. This preliminary computation allows the system to have multiple accurate reference models ready for use, improving reliability without requiring real-time generation of multiple models
Solution Approach 2:
The system captures more images than strictly necessary for a single road-height model, generating multiple models with overlapping coverage. This excessive sampling ensures that at least one model provides sufficient accuracy for reliable object position determination, with the ability to use partial models if needed
Data Source
AI summary
Systems and techniques are described herein for determining positions of objects. For instance, a method for determining positions of objects is provided. The method may include obtaining a first camera position related to a first image; obtaining a first vector, the first vector being based on the first camera position and a representation of an object in the first image; obtaining a first road-height model comprising three-dimensional positions of a road on which the object is positioned, wherein the first road-height model is related to the first camera position; obtaining a second road-height model after obtaining the first road-height model; and projecting the first vector from the first camera position to a point related to the second road-height model to determine a first object position.