DCT Image Down-Sampling Using Learning With Forgetting Algorithm

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Down-sampling of images represented in the Discrete Cosine Transform (DCT) domain is computationally complex due to the need for conversion to the spatial domain and back, which increases processing time and power consumption.

Innovation Solution

Applying two transform matrices, a row transform matrix and a column transform matrix, directly in the DCT domain to down-sample images without converting to the spatial domain, using a method that selects matrices to optimize visual quality and reduce computational complexity.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If standard down-sampling method is applied to DCT images by converting to spatial domain and back, then down-sampling can be performed, but computational complexity increases significantly

Engineering Contradiction:

Improvedown-sampling capabilityVSAvoidcomputational complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The patent replaces the conventional mechanical process of converting to spatial domain and back with a direct matrix-based transformation in the DCT domain. By substituting the complex multi-step process with a streamlined matrix multiplication approach, computational complexity is reduced while maintaining down-sampling functionality.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent changes the fundamental parameters of the down-sampling process by operating directly in the DCT domain rather than converting to spatial domain. This parameter change allows the system to perform down-sampling through matrix multiplication with optimized transform matrices, significantly reducing computational complexity while preserving image quality.

Inventive Principle:

Principle #35Parameter changes

2Productivity

If conversion to spatial domain and back is performed for down-sampling, then down-sampling can be achieved, but processing time increases

Engineering Contradiction:

Improvedown-sampling operationVSAvoidprocessing time

Core Design Contradiction:

ProductivityVSLoss of time

Solution Approach 1:

The patent performs preliminary action by pre-computing and storing optimized transform matrices that enable direct DCT domain down-sampling. These pre-computed matrices eliminate the need for time-consuming spatial domain conversion during actual down-sampling operations, significantly reducing processing time.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent substitutes the time-consuming mechanical process of inverse DCT followed by spatial down-sampling and then DCT with a direct matrix multiplication approach in the DCT domain. This substitution dramatically reduces processing time while achieving the same down-sampling effect.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Productivity

If conversion to spatial domain and back is performed for down-sampling, then down-sampling can be achieved, but power consumption increases

Engineering Contradiction:

Improvedown-sampling operationVSAvoidprocessor power consumption

Core Design Contradiction:

ProductivityVSUse of energy by stationary object

Solution Approach 1:

The patent replaces the energy-intensive mechanical process of converting to spatial domain and back with a direct matrix-based transformation in the DCT domain. By substituting multiple complex operations with a single optimized matrix multiplication, power consumption is significantly reduced while maintaining down-sampling capability.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent changes the operational parameters by performing down-sampling directly in the DCT domain rather than converting to spatial domain. This parameter change reduces the number of computational operations required, thereby reducing processor power consumption while achieving the same functional result.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS8903168B2Method and device for selecting transform matrices for down-sampling DCT image using learning with forgetting algorithm

Publication Date: 2014.12.02 MALIKIE INNOVATIONS LTD

AI summary

Down-sampling of an image may be performed in the DCT domain. A multiple layered network is used to select transform matrices for down-sampling a DCT image of size M×N to a DCT image of size I×J. A spatial domain down-sampling method is selected and applied to the DCT image to produce a down-sampled DCT reference image. A learning with forgetting algorithm is used to apply a decay to the elements of the transform matrix and select a transform matrices which solve an optimization problem. The optimization problem is a function of the visual quality of images obtained using the transform matrices and the computational complexity associated with using the transform matrices. The visual quality is a measure of the difference between the down-sampled DCT image obtained using the transform matrices and the visual quality of the DCT reference image obtained using a spatial domain down-sampling method.