Image Deblurring Using Inertially Stabilized Artificial Reference Point
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current deblurring techniques, particularly post-processing methods, often fail to accurately account for motion blur caused by relative movement between imaging systems and scenes, leading to inadequate reduction in image blur, especially when natural reference points like stars are not available or visible.
Innovation Solution
The method involves using an inertially stabilized artificial reference point object to generate a blur model based on a sequence of reference images, which is then used to deconvolve the image and reduce blur effectively, without the need for extensive inertial stabilization of the entire imaging system over long periods.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional post-processing deblurring techniques are used, then the image processing can be performed after image generation, but the motion blur caused by relative movement between imaging system and scene cannot be accurately accounted for
Solution Approach 1:
The patent applies preliminary action by capturing reference images of a stable artificial reference object before the main image acquisition. These reference images are used to pre-determine the point spread function and motion blur characteristics, which are then applied to deconvolve the main image. This allows accurate motion blur modeling without requiring continuous inertial stabilization during the entire imaging process.
Solution Approach 2:
The patent introduces an artificial reference object as an intermediary element. This reference object serves as a mediator to capture and characterize the motion blur effects during exposure. By using this intermediary, the system can accurately model the point spread function and apply it to deconvolve the main image, overcoming the limitation of traditional methods that cannot accurately account for motion blur.
2Measurement precision
If inertial stabilization is applied to the entire imaging system over long periods, then motion blur can be reduced, but the weight, size, and cost of the system increase
Solution Approach 1:
The patent extracts the inertial stabilization requirement from the main imaging system. Instead of stabilizing the entire imaging system over long periods, the method extracts only the essential stabilization need by using a separately captured reference image of a stable artificial object. This allows motion blur characterization without requiring continuous stabilization of the main imaging system, thereby reducing system complexity, weight, and cost.
Solution Approach 2:
The patent uses a copy approach by capturing a reference image of an artificial reference object that replicates the motion blur conditions. This reference copy allows the system to model motion blur effects without requiring the main imaging system to be inertially stabilized during the actual image capture. The reference image serves as a template for deconvolution, achieving motion blur reduction without the complexity of continuous stabilization.
3Productivity
If pre-processing hardware-based methods are used, then deblurring can be performed before image generation, but the cost for deblurring images increases more than desired
Solution Approach 1:
The patent uses a software-based copying approach by creating a digital copy of the point spread function from reference images. This digital copy is then applied through computational deconvolution algorithms to restore the image. This approach replaces expensive hardware-based pre-processing systems with a cost-effective software solution that achieves similar deblurring functionality, making the system more economical while maintaining processing efficiency.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach effectively reduces image blur by accurately modeling motion blur, resulting in clearer images with reduced weight, size, and cost compared to traditional systems, as the inertial stabilization is only required for the duration of a single exposure time.
Implementation Method 1
A source of the artificial reference point object is inertially stabilized over the exposure time
Implementation Method 2
The image is deconvolved with the blur model identified for the image to form a modified image having a desired reduction in blur
Data Source
AI summary
A method and apparatus for reducing blur in an image generated by an imaging system is provided. A blur model for the image is identified with respect to an exposure time for the image using a sequence of reference images of an artificial reference point object generated during the exposure time for the image. A source of the artificial reference point object is inertially stabilized over the exposure time. The image is deconvolved with the blur model identified for the image to form a modified image having a desired reduction in blur relative to the image generated by the imaging system.