Event Camera Bracketing for Motion-Robust Composite Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image bracketing techniques are time-consuming, resource-intensive, and prone to blur and misalignment due to motion during image acquisition, especially when capturing multiple images with different camera settings.
Innovation Solution
An apparatus and method utilizing an event camera and an image camera to rapidly change camera settings, such as focus distance and aperture, by recording bracketed events while maintaining fixed settings for image capture, and combining these with acquired images to produce composite images with superior properties.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If multiple images are captured with different camera settings using conventional image bracketing, then various exposure and focus options are obtained, but the acquisition process becomes time-consuming and resource-intensive
Solution Approach 1:
The patent segments the image capture process into two distinct components: an event camera that continuously records temporal changes without exposure delays, and a frame camera that captures static images at specific settings. This segmentation allows parallel operation where the event camera runs continuously while the frame camera takes discrete shots, eliminating the sequential time penalty of conventional bracketing.
Solution Approach 2:
The event camera serves as an intermediary device that bridges the gap between continuous scene monitoring and discrete image capture. It provides real-time temporal information that guides when the frame camera should capture images, enabling rapid bracketing without the frame camera needing to continuously expose and process full images for every setting change.
2Adaptability or versatility
If multiple images are captured with different camera settings, then comprehensive scene coverage is achieved, but memory and energy resources are consumed
Solution Approach 1:
The patent extracts only the essential temporal information from the continuous scene using the event camera, which records only changes in luminance rather than capturing complete image frames. This extraction approach captures all necessary dynamic scene information while consuming minimal energy and memory, as event data is highly compressed compared to full image frames.
Solution Approach 2:
The system changes the operational parameters of the two cameras differently: the event camera operates in continuous high-temporal-resolution mode with low energy consumption, while the frame camera operates in discrete capture mode only when needed. This differential parameter configuration optimizes the overall energy efficiency while maintaining comprehensive scene coverage.
3Adaptability or versatility
If multiple images are captured with different camera settings, then comprehensive focus range is obtained, but image quality degrades due to motion blur and misalignment
Solution Approach 1:
The event camera performs preliminary continuous recording of the scene before, during, and after the frame camera captures images at different focus settings. This preliminary action provides a temporal reference framework that enables precise alignment of the bracketed images, as the event data captures the exact timing and nature of any scene motion that occurred during the rapid focus transitions.
Solution Approach 2:
The system uses feedback from the event camera's continuous monitoring to detect and compensate for scene motion between frame camera exposures. By comparing event data across the bracketing sequence, the system can identify motion patterns and apply corrective transformations to align the focus-bracketed images, maintaining high precision even during object motion.
Data Source
AI summary
An apparatus and method for event-based image bracketing are provided. The apparatus includes an image camera and an event camera. The image camera acquires one or more images with fixed camera settings. The settings of the image camera are changed rapidly and events are recorded by the event camera. The acquired one or more images and part of the bracketed events are then fused to produce a composite image. A composite image can be an image with a desired camera setting, or an image with properties that cannot be captured with the image camera regardless of settings, such as an all-in-focus image. A distance of the sharpest focus setting can also be determined from the bracketed events. The focus of the image camera can then be set to this determined distance.