Image-Space Trajectory Planning for Uncertain Depth Regions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in navigating through environments with uncertain depth estimates, particularly due to complex shapes of objects like trees, which can lead to unreliable motion planning.
Innovation Solution
The implementation of image space motion planning techniques, which process captured images to identify regions with uncertain depth estimates and optimize the 3D trajectory of the autonomous vehicle to avoid these areas, thereby minimizing the risk of collisions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If visual odometry is used to estimate position and orientation based on captured images, then the autonomous vehicle can navigate using image data, but the depth estimates become unreliable in regions with complex object shapes like trees
Solution Approach 1:
The patent segments the navigation space into discrete image pixels, where each pixel represents a potential navigation target. By processing images to identify valid versus invalid depth estimate regions at the pixel level, the system can selectively navigate toward pixels with reliable depth information while avoiding pixels with uncertain depth estimates, thus resolving the contradiction between measurement precision and motion planning reliability
Solution Approach 2:
The patent introduces an intermediary processing step that analyzes image data to generate depth confidence maps before motion planning occurs. This intermediary layer identifies regions with unreliable depth estimates (such as those caused by complex object shapes) and prevents the motion planner from selecting navigation targets in these regions, thereby maintaining reliability without sacrificing the ability to use visual odometry
2Reliability
If the autonomous vehicle avoids regions with uncertain depth estimates, then collision risk is reduced, but the navigation path may be constrained to fewer options
Solution Approach 1:
The patent dynamically adjusts the set of available navigation pixels based on real-time analysis of depth estimate reliability. As the autonomous vehicle moves and captures new images, the system continuously updates which pixels represent valid navigation targets versus those with uncertain depth. This dynamic adaptation allows the vehicle to exploit available safe pathways while maintaining the ability to respond to changing environmental conditions, thus preserving versatility while ensuring safety
Data Source
AI summary
An autonomous vehicle that is equipped with image capture devices can use information gathered from the image capture devices to plan a future three-dimensional (3D) trajectory through a physical environment. To this end, a technique is described for image-space based motion planning. In an embodiment, a planned 3D trajectory is projected into an image-space of an image captured by the autonomous vehicle. The planned 3D trajectory is then optimized according to a cost function derived from information (e.g., depth estimates) in the captured image. The cost function associates higher cost values with identified regions of the captured image that are associated with areas of the physical environment into which travel is risky or otherwise undesirable. The autonomous vehicle is thereby encouraged to avoid these areas while satisfying other motion planning objectives.