Iterative Back Projection Signal Processing for Image Detail Enhancement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing super-resolution imaging technologies require significant computation and enhance high-frequency noise, making them impractical for practical applications and failing to effectively enhance image details while reducing noise.
Innovation Solution
A signal processing device and method that utilizes a processor to perform iterative back projection and three-dimensional noise reduction, selectively adding super resolution difference and noise reduction values to frames stored in buffers to generate an output frame, reducing hardware and software storage costs and noise.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If deep learning based super-resolution imaging is used, then image details are enhanced, but computation requirements become huge
Solution Approach 1:
The patent replaces the deep learning computational system with a signal processing system based on iterative back projection and three-dimensional noise reduction. This substitution uses mathematical operations and signal processing techniques instead of complex neural network computations, significantly reducing the computational power required while maintaining image enhancement capabilities
Solution Approach 2:
The patent uses frame buffering and iterative processing where previous frames are stored and reused as reference data. By copying and reusing historical frame data in the buffer, the system avoids redundant computations and reduces the overall computational burden while maintaining consistent image quality across frames
2Manufacturing precision
If super-resolution imaging is performed, then image details are enhanced, but high-frequency noise is also enhanced
Solution Approach 1:
The patent converts the harmful high-frequency noise into a beneficial filtering process. By applying three-dimensional noise reduction in the temporal and spatial domains, the system identifies and removes noise components while preserving genuine image details. The iterative back projection process further refines this by projecting and reconstructing image data to distinguish between noise and actual image features
Solution Approach 2:
The patent applies different processing treatments to different regions and frequency components of the image. The noise reduction process selectively targets high-frequency components while preserving low-frequency image details. The iterative back projection also applies local adjustments based on the specific characteristics of each region, enhancing details where needed while suppressing noise in other areas
3Object-generated harmful factors
If noise reduction is performed in time domain and space domain, then high-frequency noise is reduced, but storage cost increases
Solution Approach 1:
The patent segments the noise reduction process into distinct temporal and spatial domains. By separating the processing into these two domains, the system can efficiently manage data storage requirements. The frame buffer stores only essential historical frame data needed for temporal noise reduction, while spatial noise reduction operates on the current frame and its neighbors, optimizing the balance between noise reduction effectiveness and storage cost
Data Source
AI summary
A signal processing device includes a first frame buffer configured to store a first frame, a second frame buffer configured to store a second frame and a processor. The processor is coupled to the first frame buffer and the second frame buffer and is configured to perform a first image processing procedure according to the first frame and the second frame to obtain a super resolution difference value corresponding to each pixel of the first frame, perform a second image processing procedure according to the first frame and the second frame to obtain a noise reduction value corresponding to each pixel of the first frame, selectively add the super resolution difference value and the noise reduction value to the corresponding pixel of the first frame to generate an output frame and store the output frame in the second frame buffer as the second frame.