Long Exposure Image Simulation via Motion Compensation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Camera shake during long exposure times in digital photography results in motion blur and ghost images, particularly in low light conditions, where a tripod is inconvenient to use.
Innovation Solution
Capturing three images at short exposure times and using block matching and minimum square error algorithms to remove relative motion, combining them to simulate an image with long exposure without motion blur.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Illumination intensity
If a longer shutter speed is used to fully expose the image in low light, then the image exposure is improved, but camera shake causes motion blur and ghost images
Solution Approach 1:
The patent segments the long exposure capture into multiple short exposure images (first, second, and third images). Each image is captured with a short exposure time to avoid motion blur, then the images are processed individually through block matching and transformation before being combined to create the final simulated long exposure image.
Solution Approach 2:
The patent performs preliminary actions by capturing multiple short exposure images before combining them. The block matching and transformation processes are performed in advance on each individual image to remove relative motion, preparing them for the final combination step that simulates the long exposure effect.
2Reliability
If a tripod is used to stabilize the camera, then camera shake is eliminated, but the device complexity and convenience are worsened
Solution Approach 1:
The patent replaces the mechanical stabilization system (tripod) with a computational approach. Instead of physically stabilizing the camera, the system captures multiple short exposure images and uses block matching algorithms and geometric transformations to computationally remove the effects of camera shake, achieving stabilization without additional mechanical equipment.
Solution Approach 2:
The patent creates multiple copies of the scene captured at different moments (three separate short exposure images). These copies are then processed and combined to produce a final image that represents what would have been captured with a long exposure, eliminating the need for physical stabilization equipment.
3Reliability
If multiple images are captured and combined to simulate long exposure, then motion blur is eliminated, but the processing complexity increases
Solution Approach 1:
The processing is segmented into distinct stages: block matching to find corresponding points between images, RANSAC-based transformation to remove relative motion, and final combination of processed images. This segmentation of the processing pipeline manages complexity by breaking down the overall task into manageable, specialized sub-tasks.
Solution Approach 2:
The patent uses an excessive number of images (three images) for the simulation process, which provides redundancy that helps in achieving better motion compensation and more accurate long exposure simulation. This partial/excessive action approach ensures robustness in the processing despite the increased computational requirements.
Data Source
AI summary
A method for simulating an image captured at a long exposure time (“simulated image”), includes (1) capturing each of first, second, and third images at a short exposure time, (2) determining a first relative motion between the first and the second images, (3) transforming the first image to remove the first relative motion, (4) determining a second relative motion between the third and the second images, (5) transforming the third image to remove the second relative motion, and (6) combining the first, the second, and the third images to form the simulated image. Relative motions between images are determined by matching blocks at multiple resolutions to determine corresponding points between the images. Transformation to remove relative motion is determined by fitting corresponding points between the images using a minimum square error (MSE) algorithm in a random sample consensus (RANSAC) framework.