Image Block Processing Using Modulo Pixel Calculation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing technologies face inefficiencies in dividing images into blocks for parallel processing, particularly when the image size is not an integer multiple of the block size, leading to increased processing time and overhead due to the need for adjusting image sizes and adding dummy pixels.
Innovation Solution
A method that determines the starting end of an image block using a formula (N = (n-1)Wa + (m-1)Wb + 1) to obtain image data in chunks corresponding to the block size, allowing efficient processing regardless of image size, without requiring dummy pixels or adjustments based on block size relations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If dummy pixels are added to adjust image size to integer multiple of block size, then image blocks can be divided uniformly, but processing time increases due to overhead and size adjustment operations
Solution Approach 1:
The patent extracts and eliminates the dummy pixel addition step from the conventional image processing workflow. By using modulo arithmetic to calculate the actual number of pixels needed (Wa mod Wb), the system processes only the necessary image pixels without adding unnecessary dummy pixels, thereby removing the source of processing time overhead while maintaining proper block division uniformity.
Solution Approach 2:
The patent changes the parameter approach from fixed integer multiple adjustment to flexible modulo-based calculation. Instead of forcing image width to be an integer multiple of block width by adding dummy pixels, the system uses parameter calculation (Wa mod Wb) to determine the exact number of pixels needed, allowing efficient processing without time-consuming size adjustments.
2Productivity
If image size is adjusted to match block size relations, then parallel processing efficiency can be optimized, but device complexity increases due to conditional process changes
Solution Approach 1:
The patent creates a universal processing formula that works for all image sizes without requiring different processing paths. The modulo arithmetic calculation (Wa mod Wb) provides a single unified approach that determines pixel count for any image width, eliminating the need for multiple conditional processes and reducing device complexity while maintaining parallel processing efficiency.
Solution Approach 2:
The patent uses parameter changes through modulo arithmetic to create a universal solution. By calculating the number of pixels as (Wa mod Wb), the system adapts to any image size without requiring complex conditional logic, thereby reducing process complexity while preserving the ability to optimize parallel processing efficiency across different image dimensions.
3Ease of operation
If sequential processing is used for each pixel, then processing simplicity is maintained, but arithmetic operation amount becomes huge and processing efficiency decreases
Solution Approach 1:
The patent applies segmentation by dividing the image into blocks of Wb pixels width. This segmentation allows parallel processing of multiple pixels simultaneously while maintaining simple processing logic for each block. The modulo-based pixel calculation ensures proper segmentation boundaries without complex operations, thereby improving processing efficiency while keeping the approach simple and easy to implement.
Data Source
AI summary
A storage section configured to store image data of a raster form therein, a processing section configured to process the image data in a unit of an image block having a predetermined block size in a raster direction, and a control section configured to provide the processing section with information related to the block size and an image size are provided, and the processing section performs counting from the beginning of the image data, and obtains, in a lump, from the storage section, image data corresponding to Wb pixels that are successive from an N-th pixel represented by N=(n−1)Wa+(m−1)Wb+1, wherein Wa and Wb respectively denote the image size and the block size in the raster direction, m and n denote parameters for specifying the location of an image block.