Hyper-resolution imaging via co-registered multi-camera deconvolution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing super-resolution imaging techniques face challenges in achieving stable and efficient image registration from motion data, particularly due to the ill-posed nature of the inverse problem and the instability of blind deconvolution processes, which complicates the reconstruction of high-resolution images from low-resolution data, especially in real-world applications with complex blurring conditions.
Innovation Solution
The method involves co-registering multiple sequences of images from cameras with distinct spatial, temporal, and spectral resolutions, using a computer program product that includes software modules for storing and deconvolving images to achieve a resultant sequence with spatial resolution exceeding the original camera imaging resolution, employing Tikhonov regularization and passive image registration based on motion-derived kinetic point spread functions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If blind deconvolution is used to reconstruct high-resolution images from low-resolution data, then spatial resolution can be improved, but the process becomes unstable and computationally complex
Solution Approach 1:
The patent applies Tikhonov regularization as a preliminary constraint before performing deconvolution. By pre-defining the point spread function (PSF) based on motion data and applying regularization constraints, the method stabilizes the inverse problem solution before reconstruction, preventing the instability that would otherwise occur during blind deconvolution
Solution Approach 2:
The patent introduces motion-derived kinetic PSF as an intermediary between the low-resolution input images and the final high-resolution reconstruction. This intermediary element (the PSF) acts as a known constraint that mediates the deconvolution process, replacing the need for unstable blind deconvolution while still achieving super-resolution
2Measurement precision
If multiple camera systems with different resolutions are co-registered, then hyper-resolution imaging can be achieved, but device complexity increases
Solution Approach 1:
The patent merges multiple camera systems (e.g., EO and IR cameras) into a unified hybrid imaging system with a common coordinate framework. By co-registering the cameras and combining their data streams through a single processing pipeline that uses motion-derived PSF for all sensors, the system achieves hyper-resolution without proportionally increasing operational complexity
Solution Approach 2:
The patent creates a universal processing framework that handles multiple camera types and resolutions through a single deconvolution algorithm. The motion-derived kinetic PSF approach serves as a universal solution that works across different sensor modalities (visible light, infrared, etc.), allowing one system to perform multiple imaging functions simultaneously
3Measurement precision
If passive image registration based on motion data is used, then sub-pixel shift precision is improved, but the difficulty of detecting and measuring motion increases
Solution Approach 1:
The patent makes the imaging system self-sufficient by deriving its own motion information directly from the image sequences without external sensors. The system performs passive image registration using only the captured images and their temporal changes, allowing it to self-determine sub-pixel shifts and generate its own kinetic PSF without requiring additional measurement equipment
Data Source
AI summary
Methods and a computer program product for deriving a super-resolution image of a physical object by fusing cameras of multiple resolutions (spatial, temporal, or spectral), the super-resolution image characterized by a resolution exceeding a “camera imaging resolution” associated with each of a sequence of lower-resolution images of the physical object. The sequence of images of the physical object is obtained at a plurality of relative displacements with respect to the object by a hybrid camera system comprising at least two imaging systems. The imaging systems are characterized by respective temporal and spatial resolution and by spectral sensitivity, and may be distinct from one another in one or more of the foregoing dimensions. The imaging systems are either fixed, or subject to know motion, relative to each other. Image sequences derived by each imaging system are coregistered and deconvolved to solve for a resultant sequence of images.