Sub-aperture Image Extraction Using Blind Signal Separation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing processing methods for obtaining sub-aperture images from raw data of a plenoptic camera are not accurate due to misalignment between micro-images and sensor arrays, leading to suboptimal extraction of sub-aperture images.
Innovation Solution
A blind signal separation technique is applied using a mixing matrix derived from pixel positions, allowing for the recovery of sub-aperture images by accounting for subpixel positions of micro-image centers and maintaining consistent interpolations with physics.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the conventional processing method is used to obtain sub-aperture images from raw data, then the processing is simple and fast, but the accuracy is reduced due to misalignment between micro-images and sensor arrays
Solution Approach 1:
The patent applies preliminary action by performing a calibration process before the actual sub-aperture image extraction. The calibration step pre-determines the precise positions of micro-image centers and computes correction coefficients that are then used in the extraction process. This preliminary preparation eliminates misalignment issues during the main processing, thereby improving accuracy without significantly increasing the complexity of the primary extraction operation.
Solution Approach 2:
The patent employs parameter changes by introducing correction coefficients that adjust the mapping between sensor pixels and micro-image centers. Instead of using a simple integer-based mapping, the method modifies the parameters by incorporating sub-pixel position information and computing correction factors. This parameter adjustment resolves the misalignment problem and improves extraction accuracy while maintaining computational efficiency.
2Reliability
If the hypothesis of perfect match between micro-image centers and sensor pixels is assumed, then the processing is simplified, but the reliability is reduced due to misalignment in practice
Solution Approach 1:
The patent implements feedback by using the calibration process to determine actual micro-image center positions and comparing them with the ideal integer pixel positions. The difference (misalignment) is fed back into the system through correction coefficients that adjust the extraction process. This feedback mechanism ensures that the processing adapts to the actual physical configuration, thereby improving reliability while keeping the processing method manageable through systematic correction.
3Measurement precision
If calibration is performed to correct misalignment as proposed in US 2014/0146184, then the accuracy is improved, but the processing time and complexity increase
Solution Approach 1:
The patent applies preliminary action by performing the calibration process once during setup or initialization, and then reusing the computed correction coefficients for all subsequent sub-aperture image extractions. This preliminary calibration eliminates the need for repeated time-consuming alignment computations during normal operation, thereby achieving high accuracy while minimizing processing time loss.
Solution Approach 2:
The patent employs parameter changes by computing correction coefficients that capture the misalignment characteristics. Once these parameters are determined through calibration, they are applied directly in the extraction process without requiring repeated optimization or adjustment. This parameter-based correction approach achieves accurate results efficiently, reducing processing time compared to iterative realignment methods.
Data Source
Figure 1
Figure 2
Figure 3~4
AI summary
In one embodiment, it is proposed a method for obtaining at least one sub-aperture image being associated with one view, from raw light field data corresponding to recorded data by an array of pixels sensors positioned behind an array of micro-lenses in a light field camera, each of said pixel sensor recording a linear mixing of up to four different views. The method is remarkable in that it comprises applying a signal separation process on said raw data by using an inverse of a mixing matrix A, said mixing matrix comprising coefficients that convey weighting information of said up to four different views recorded by a pixel sensor.