Autonomous Vehicle Pose Estimation From Single-Image Motion Blur
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in accurately determining pose changes due to high computational complexity and reliance on multiple images, which can be hindered by environmental factors like camera obstructions, leading to localization errors.
Innovation Solution
Implementing a machine learning model that processes single images to determine pose changes based on blurring, using sensors to capture images while moving, and aggregating pose data from multiple sensors to improve localization accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple images are used for pose estimation, then localization accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent extracts and utilizes motion blur information from individual images as a separate feature source. Instead of processing multiple complete images, the system extracts blur characteristics (kernel size, orientation, intensity) from single images to estimate pose changes, thereby reducing computational complexity while maintaining localization accuracy
Solution Approach 2:
The patent introduces motion blur as an intermediary element that carries pose information. The blur kernel acts as a mediator between the moving camera and the static scene, encoding motion characteristics that can be decoded to estimate pose changes without requiring multiple images
2Measurement precision
If multiple images are processed simultaneously, then pose estimation accuracy is improved, but processing speed decreases
Solution Approach 1:
The patent segments the pose estimation task into independent single-image processing units. Each image is processed individually to extract blur features and estimate pose changes, allowing parallel processing and real-time operation without the computational burden of processing multiple images simultaneously
3Reliability
If the camera is blocked by objects or environmental factors, then image quality deteriorates, but localization may still be performed using blur information
Solution Approach 1:
The patent converts the harmful effect of motion blur (caused by camera obstruction or vehicle motion) into a beneficial feature for pose estimation. Instead of treating blur as noise to be removed, the system uses blur characteristics as informative signals to estimate pose changes, making the system robust to obstructions and environmental factors
Data Source
AI summary
A vehicle comprises a sensor configured to capture images and one or more processors. The one or more processors can be configured to receive a single image from the sensor, the single image captured by the sensor as the autonomous vehicle was moving; execute a machine learning model using the single image as input to generate a change in pose of the autonomous vehicle, the machine learning model trained to output changes in pose of autonomous vehicles based on blurring in individual images; determine a global position of the autonomous vehicle based on the generated change in pose of the autonomous vehicle; and transmit the global position to an autonomous vehicle controller configured to control the autonomous vehicle.