Image Metadata Generation via Frequency Band Segmentation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Display devices face challenges in efficiently processing images with many edges, as they require significant hardware resources to accurately analyze frequency characteristics, which affects image sharpness and realism.

Innovation Solution

A method and device for generating metadata that includes frequency characteristic information of an image by using discrete cosine transform (DCT) coefficients, dividing blocks into frequency bands, and determining band values based on threshold comparisons, allowing for clustering into high-frequency, mid-frequency, or low-frequency images.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If display devices process images with many edges to accurately analyze frequency characteristics, then image sharpness and realism are improved, but hardware resource consumption increases significantly

Engineering Contradiction:
Improvefrequency characteristic analysis accuracyVSAvoidhardware resource consumption
Core Design Contradiction:
Measurement precisionVSUse of energy by moving object

Solution Approach 1:

The image is divided into multiple blocks, and each block is further divided into frequency bands (low-frequency, mid-frequency, high-frequency regions). This segmentation allows the device to analyze frequency characteristics of individual blocks rather than processing the entire image uniformly, reducing overall hardware resource consumption while maintaining analysis accuracy for edge-rich regions.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent applies different processing strategies to different regions of the image based on their frequency characteristics. Blocks with high-frequency characteristics (containing edges) receive more detailed analysis, while low-frequency blocks use simpler processing. This local quality approach ensures accurate frequency analysis where needed while conserving hardware resources in regions requiring less processing.

Inventive Principle:
Principle #3Local quality

2Measurement precision

If display devices use more hardware resources to process images with many edges, then image sharpness and realism are improved, but processing efficiency decreases

Engineering Contradiction:
Improvefrequency characteristic analysis accuracyVSAvoidimage processing efficiency
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The patent applies full frequency analysis only to blocks that require it (those with high-frequency characteristics), while using simplified processing for other blocks. This partial action approach maintains measurement precision for critical regions while improving overall processing efficiency by avoiding unnecessary detailed analysis in regions where it is not needed.

Inventive Principle:
Principle #16Partial or excessive action

Solution Approach 2:

By segmenting the image into blocks and analyzing their frequency characteristics separately, the device can identify and focus computational resources only on blocks containing edges or high-frequency content. This segmentation strategy improves processing efficiency by avoiding uniform high-cost processing across the entire image while maintaining accuracy where required.

Inventive Principle:
Principle #1Segmentation

3Measurement precision

If the image is divided into multiple frequency bands for detailed analysis, then frequency characteristic accuracy is improved, but device complexity increases

Engineering Contradiction:
Improvefrequency characteristic information accuracyVSAvoidprocessing structure complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent divides each image block into three frequency bands (low-frequency, mid-frequency, high-frequency regions) and analyzes them separately. This segmentation provides detailed frequency characteristic information while keeping the processing structure manageable by using a consistent three-band division across all blocks, avoiding the need for complex variable-band processing.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent determines frequency characteristics by comparing statistics of frequency coefficients against threshold values for each band. This parameter-based approach (using threshold comparisons) provides accurate frequency classification while maintaining relatively simple device structure, avoiding the need for complex machine learning models or sophisticated signal processing algorithms.

Inventive Principle:
Principle #35Parameter changes

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

This approach enables more efficient use of hardware resources and improved image realism by accurately analyzing and transmitting frequency characteristics, allowing for better image processing and display, particularly in UHD displays.

Implementation Method 1

The frequency coefficients may include discrete cosine transform (DCT) coefficients generated by performing DCT on the pixel values

Methodology Applied
Scientific EffectDiscrete Cosine Transform:

Data Source

PatentEP3211887B1Method and device for generating metadata including frequency characteristic information of image
Publication Date: 2020.07.01 SAMSUNG ELECTRONICS CO LTD
  • EP3211887B1 patent drawingFigure 1~2
  • EP3211887B1 patent drawingFigure 3
  • EP3211887B1 patent drawingFigure 4

AI summary

Provided are a method and device for generating metadata including frequency characteristic information of an image. Pixel values of a current block among blocks divided from the image are converted into frequency coefficients in the frequency domain. A band value of a frequency band corresponding to each of regions of the current block is determined using the frequency coefficients included in the regions of the current block, the regions of the current block being divided to correspond to different frequency bands. Metadata including the frequency characteristic information of the current block is generated based on the determined band values.