Multi-Sensor HDR Image Alignment and Blending
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional high dynamic range (HDR) imaging systems introduce motion blur when capturing moving objects due to the sequential generation of images at different exposures, making object tracking and event detection difficult in environments with varying lighting conditions.
Innovation Solution
A multiple-image sensor system that generates image data simultaneously at different exposures, allowing for real-time alignment and blending of images from overlapping fields of view without motion blur, using preprocessing, matching, alignment, and blending processes to create HDR images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Illumination intensity
If sequential image capture at different exposures is used, then high dynamic range imaging is achieved, but motion blur is introduced
Solution Approach 1:
The patent divides the image capture function into multiple parallel image sensors, each capturing at a different exposure level simultaneously. This segmentation allows the system to capture multiple exposure values without sequential timing, thereby eliminating motion blur while maintaining high dynamic range capability.
Solution Approach 2:
The patent transitions from temporal dimension (sequential capture) to spatial dimension (parallel sensors with overlapping fields of view). By adding spatial separation through multiple sensors positioned at different locations, the system achieves high dynamic range without the temporal sequence that causes motion blur.
2Manufacturing precision
If multiple image sensors with overlapping fields of view are used, then motion blur is eliminated, but device complexity increases
Solution Approach 1:
The patent merges multiple image sensors into a coordinated system where sensors with overlapping fields of view work together. By combining the data from multiple sensors and using image blending techniques, the system achieves high dynamic range imaging without motion blur, while the overlapping fields of view provide redundancy that simplifies alignment requirements.
Solution Approach 2:
The patent changes the exposure parameter across multiple sensors simultaneously, with each sensor capturing at a different exposure level. This parameter variation allows the system to capture the full dynamic range in a single time instant, eliminating motion blur while the automated processing algorithms manage the complexity of handling multiple sensors.
3Measurement precision
If image blending from multiple sensors is performed, then accurate detection in varying lighting is achieved, but processing complexity increases
Solution Approach 1:
The patent implements self-service through automated image processing algorithms that automatically align, match, and blend images from multiple sensors. The system uses feature detection and homography estimation to autonomously handle the complex processing tasks, reducing manual intervention while achieving accurate detection in varying lighting conditions.
Solution Approach 2:
The patent employs feedback mechanisms in the image processing pipeline, where the system continuously refines alignment and blending based on detected features and image quality metrics. This feedback loop ensures accurate detection by adjusting processing parameters based on the actual image data, improving measurement precision while managing processing complexity through iterative optimization.
Data Source
AI summary
Described are systems and methods for generating high dynamic range (“HDR”) images based on image data obtained from different image sensors for use in detecting events and monitoring inventory within a materials handling facility. The different image sensors may be aligned and calibrated and the image data from the sensors may be generated at approximately the same time but at different exposures. The image data may then be preprocessed, matched, aligned, and blended to produce an HDR image that does not include overexposed regions or underexposed regions.