Motion Blur Feature Extraction for Indoor Robot Localization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Indoor navigation systems for robots, such as drones, face challenges in accurately determining position due to signal attenuation and multi-path effects in indoor environments, and existing optical localization systems struggle with varying lighting conditions and feature extraction, leading to errors and reduced robustness.
Innovation Solution
A method for identifying candidate features in images using spatial variations in image sharpness, determined through deep neural networks or classical methods, to enhance feature extraction and reduce errors in indoor navigation systems, involving controlled camera and light source movements to induce specific blurriness and sharpness in captured images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If optical localization systems extract features from images in varying lighting conditions, then localization can be performed, but feature extraction becomes unreliable due to illuminance variations and dynamic range limitations
Solution Approach 1:
The patent applies parameter changes by deliberately modifying the camera exposure time parameter to create motion blur. By controlling the exposure duration, the system transforms the image to have reduced sharpness while preserving feature information, making features more detectable against varying lighting conditions and improving extraction reliability.
Solution Approach 2:
The patent converts the harmful effect of motion blur (which typically degrades image quality) into a beneficial feature extraction aid. By introducing controlled motion blur through extended exposure, the system enhances feature detectability and robustness, turning a normally detrimental artifact into a useful transformation for overcoming lighting variability.
2Illumination intensity
If camera exposure time is extended to improve low-light visibility, then image brightness improves, but motion blur increases and sharpness decreases
Solution Approach 1:
The patent deliberately changes the exposure time parameter to an extended duration, which simultaneously increases image brightness for low-light conditions and introduces motion blur. This parameter change is intentional and controlled, transforming the brightness-sharpness trade-off into a beneficial state for feature extraction in challenging lighting.
Solution Approach 2:
The patent converts the harmful motion blur effect into a beneficial feature for robust feature extraction. By accepting and utilizing the blur introduced by extended exposure, the system achieves better feature detectability and localization accuracy despite the loss of traditional image sharpness.
3Measurement precision
If features are extracted from blurry images, then feature detection becomes more difficult, but in this invention the blur is controlled to preserve feature information for improved localization
Solution Approach 1:
The patent changes the image parameters through controlled motion blur to preserve feature information while reducing the impact of lighting variations. This parameter transformation makes features more detectable and robust, improving measurement precision despite the intentional blurring.
Solution Approach 2:
The system uses feedback from the blurred image to identify and locate features, then uses this information to determine camera pose and update the localization model. The feedback loop allows the system to adapt to the blurred image characteristics and maintain accurate feature detection and localization.
Data Source
AI summary
The present invention relates to a method for identifying at least one candidate feature in an image of a scene of interest captured by a camera, and to a method for capturing an image, with spatial variations in image sharpness, of a scene of interest by a camera, and to a method for determining a state xk of a camera at a time tk, as well as to an assembly and two computer program products.