Image Processing Apparatus Superpixel Graph Cut Area Detection

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional graph cut area division methods in image processing fail to consider texture information from collections of pixels, leading to inaccurate area detection.

Innovation Solution

An image processing apparatus and method that divides images into small areas based on edge information, estimates attribute probabilities for each area, calculates connection strengths between adjacent areas, and uses these probabilities and strengths to detect specific areas using a graph cut method, considering features from multiple pixels.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If area division is performed using conventional graph cut method with each pixel as a node, then the method can process images systematically, but it cannot consider texture information from collections of pixels leading to inaccurate area detection

Engineering Contradiction:
Improvearea detection accuracyVSAvoidtexture information
Core Design Contradiction:
Measurement precisionVSLoss of information

Solution Approach 1:

The patent divides the image into multiple small areas (super-pixels) grouped from adjacent pixels before applying graph cut method. This segmentation allows texture information from collections of pixels to be considered while maintaining systematic processing capability.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent merges adjacent pixels into small areas (super-pixels) to preserve texture information. By combining multiple pixels into unified regions, the method retains collective pixel features that characterize texture patterns.

Inventive Principle:
Principle #5Merging (Combining)

2Reliability

If area division is performed by creating a connected graph with multiple nodes corresponding to multiple pixels, then the graph structure can represent pixel relationships, but the arithmetic operation becomes complex and computationally intensive

Engineering Contradiction:
Improvearea division accuracyVSAvoidcomputational complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent segments the image into small areas (super-pixels) before graph construction. This reduces the number of nodes in the connected graph from individual pixel count to super-pixel count, significantly lowering computational complexity while maintaining area division accuracy.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent creates a simplified representation of the image by copying and grouping pixel information into super-pixels. This reduced representation maintains essential visual information while decreasing the computational burden of graph operations.

Inventive Principle:
Principle #26Copying

Data Source

PatentEP2523164B1Image processing apparatus, image processing method, and image processing program
Publication Date: 2018.07.11 OLYMPUS CORPORATION(JP)
  • EP2523164B1 patent drawingFigure 1
  • EP2523164B1 patent drawingFigure 2
  • EP2523164B1 patent drawingFigure 3

AI summary

An image processing apparatus includes a small area divider (110) that divides, on the basis of edge information of an image, the image into multiple small areas each including multiple pixels; an attribute probability estimator (120, 220) that estimates attribute probability for each of the small areas, which is probability that the small area is attributed to a specific area to be detected; an adjacent-small-area connection strength calculator (130, 230, 330, 340, 430) that calculates connection strength that quantitatively indicates a degree to which small areas adjacent to each other among the multiple small areas are attributed to the same area that is the specific area or a non-specific area; and a specific area detector (140) that detects the specific area on the basis of the attribute probability and the connection strength.