Image Processing Device Scaling Gain Calculation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing techniques face challenges in up-scaling images without degrading quality, particularly due to the loss of precision in high band components when increasing sampling frequency, leading to increased memory requirements for enhancement processes.
Innovation Solution
An image processing device and method that includes a scaling unit for enlarging images, a gain calculating unit to compensate for phase deviation, and an adding unit to enhance the enlarged image by adding a calculated gain to each pixel, allowing for up-scaling and enhancement while minimizing memory requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If enhancement process is conducted after up-scaling, then image quality degradation is prevented, but memory capacity requirements increase
Solution Approach 1:
The patent applies preliminary action by calculating the gain for enhancement based on the input image signal before the up-scaling operation is completed. The gain calculating unit computes enhancement gains using the original low-resolution image data, and these pre-calculated gains are then applied to the up-scaled high-resolution image. This approach allows the enhancement process to begin before the full high-resolution image is generated, reducing the memory burden of storing complete high-resolution image data during enhancement calculations.
Solution Approach 2:
The patent segments the image processing into distinct functional units: a scaling unit for up-scaling, a gain calculating unit for computing enhancement gains, and an adding unit for applying the gains. The gain calculating unit processes the image in a segmented manner by calculating gains for different regions or pixels independently based on the input image signal, rather than requiring the entire up-scaled image to be stored and processed as a single unit. This segmentation reduces peak memory requirements.
2Ease of operation
If separate memories are used for input image and enlarged image, then enhancement process can be performed, but device complexity increases
Solution Approach 1:
The patent implements multi-functionality by designing a memory structure that serves dual purposes: it stores the input low-resolution image signal for gain calculation and simultaneously provides reference data for the enhancement process applied to the up-scaled image. The same memory resource is used for both the original image storage and as a reference for calculating enhancement gains, eliminating the need for completely separate memory systems for input and output images.
Solution Approach 2:
The patent merges the functionality of storing input images and reference data into a unified memory structure. The memory system is designed to hold the input image signal that serves both as the source for up-scaling and as the reference for calculating enhancement gains. This merging of storage functions reduces the overall memory architecture complexity compared to maintaining entirely separate memory systems for input and processed images.
3Device complexity
If up-scaling is performed without considering Nyquist frequency, then processing is simpler, but high band component precision is lost
Solution Approach 1:
The patent uses a copying approach by preserving the input image signal data as a reference copy throughout the up-scaling and enhancement process. The gain calculating unit creates a computational copy of the enhancement information based on the original input image, which is then applied to the up-scaled version. This copying of the original image characteristics into the enhancement gain calculations ensures that high-frequency information from the original image is not lost during up-scaling, while avoiding the complexity of redesigning the up-scaling filter itself.
Data Source
AI summary
Provided is an image processing device including a scaling unit configured to enlarge an input image represented by an input image signal to generate an enlarged image, a gain calculating unit configured to calculate a gain for enhancing the enlarged image for each target pixel of the enlarged image on a basis of the input image signal, and an adding unit configured to add, for each target pixel, a pixel value of the target pixel and the gain corresponding to the pixel value of the target pixel.