Integer Coefficient Table for Image Scaling Without Dividers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image scaling processes require a divider and a floating point unit, leading to complex device configurations and redundant operations due to the need for scaling up and storing image data, as well as intermediate images.
Innovation Solution
A method for creating an interpolation coefficient table that allows image scaling by arbitrary multiples without using a divider or floating point unit, involving the design of a filter with a transfer function, conversion of interpolation coefficients to integers, and sorting these coefficients for efficient storage and use in the scaling process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional image scaling process is used, then image scaling by arbitrary multiples is achieved, but device complexity increases due to requirement of divider and floating point unit
Solution Approach 1:
The patent changes the parameter type of interpolation coefficients from floating-point to integer values. By representing coefficients as integers and using fixed-point arithmetic, the system achieves arbitrary image scaling without requiring floating-point units or dividers, thus reducing device complexity while maintaining scaling versatility
Solution Approach 2:
The patent replaces the mechanical/computational system of floating-point arithmetic and division operations with a simplified system using integer multiplication and fixed-point arithmetic. This substitution eliminates the need for complex hardware components (divider and floating-point unit) while achieving the same image scaling function
2Adaptability or versatility
If image is scaled up to U times and then scaled down to 1/D times, then arbitrary scaling is achieved, but redundant operation and memory usage occur
Solution Approach 1:
The patent performs preliminary action by pre-calculating and storing integer-converted interpolation coefficients in a coefficient table during system initialization. This pre-computation eliminates the need for complex runtime calculations, allowing efficient arbitrary scaling operations without redundant processing or intermediate image storage
Solution Approach 2:
The patent segments the scaling operation into two independent components: a scaling-up factor (U) and a scaling-down factor (D). By representing the overall scaling factor as U/D and using integer coefficients, the system can perform arbitrary scaling in a single operation without requiring sequential up-scaling and down-scaling steps
3Measurement precision
If floating point unit and divider are used, then precise interpolation is achieved, but device configuration becomes complicated
Solution Approach 1:
The patent uses fixed-point integer arithmetic instead of floating-point arithmetic, which can be implemented with simpler, cheaper hardware. The integer coefficients are pre-calculated with sufficient precision and stored in lookup tables, providing accurate interpolation results without requiring expensive floating-point units or dividers
Data Source
AI summary
An image scaling processor includes a coefficient table storing interpolation coefficients, a multiplier that multiplies each of pieces of pixel data constituting input image data and each of the interpolation coefficients stored in the coefficient table, an adder that repeatedly adds pieces of multiplied data output from the multiplier and obtains a total sum of the pieces of multiplied data for a predetermined number of the pieces of pixel data, a selector that outputs a total sum of the pieces of multiplied data at a timing at which the total sum of the pieces of multiplied data is obtained for the predetermined number of the pieces of pixel data, and a shift circuit that shifts an output of the selector to adjust a bit count of the output image data to a bit count of the input image data.