Gerchberg-Saxton Phase Retrieval Padding Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for retrieving phase information from images using the Gerchberg-Saxton algorithm are inefficient, requiring a large number of iterations to achieve acceptable quality, which is not suitable for real-time applications like 2D video projection.
Innovation Solution
The modified Gerchberg-Saxton algorithm incorporates padding pixels to increase the number of pixels in the hologram plane, suppresses noise in these pixels using self-setting padding and attenuation, and optimizes parameters like gain and attenuation factors to reduce iterations and improve reconstruction quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the Gerchberg-Saxton algorithm is used to retrieve phase information from images, then phase information can be obtained for hologram reconstruction, but a large number of iterations are required which makes it unsuitable for real-time applications
Solution Approach 1:
The patent applies preliminary action by pre-padding the input image with zero values before phase retrieval. This preprocessing step prepares the data in advance to facilitate faster convergence during the iterative algorithm, reducing the number of iterations needed while maintaining phase information accuracy.
Solution Approach 2:
The patent changes parameters by modifying the Gerchberg-Saxton algorithm to include padding operations and adjusting the iterative process parameters. These parameter changes optimize the algorithm's convergence behavior, enabling real-time performance while preserving the accuracy of retrieved phase information.
2Manufacturing precision
If padding pixels are added to increase the number of pixels in the hologram plane, then reconstruction quality improves, but noise in the padding pixels must be suppressed
Solution Approach 1:
The patent converts the harmful noise in padding pixels into a beneficial element by using the padding region to absorb artifacts and improve the overall reconstruction quality. The padding pixels, which would normally be sources of noise, are instead used to enhance the hologram quality by providing additional spatial frequency information.
3Productivity
If the number of iterations is reduced for real-time processing, then processing speed increases, but reconstruction quality may deteriorate
Solution Approach 1:
By pre-padding the input image and preparing the algorithm parameters in advance, the system achieves faster convergence in fewer iterations. This preliminary preparation enables real-time processing speeds while maintaining high reconstruction quality that would otherwise require many more iterations to achieve.
Data Source
Figure 1
Figure 2
Figure 3~4
AI summary
There is provided a method of processing a pixellated image to retrieve a phase distribution representative of the image. The phase distribution representative of the image is in the Fourier domain. The method comprises padding the image pixels with padding, or non-image, pixels to increase the total number of pixels in the pixellated image. The method further comprises processing the padding or non-image pixels differently to the image pixels in each iteration of the iterative method.