Texture Energy Map Body Part Detection via Convolution

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing adult image identification techniques classify images based on overall sexuality without detecting specific regions-of-interest, limiting their application in various services.

Innovation Solution

A method and apparatus that analyze texture energy maps to identify candidate body parts using candidate masks and machine-learning models for accurate detection of specific human body parts arousing sexuality.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If overall sexuality estimation is used to classify images, then classification speed is improved, but detection precision of specific body parts deteriorates

Engineering Contradiction:

Improveclassification speedVSAvoiddetection precision of specific body parts

Core Design Contradiction:

ProductivityVSMeasurement precision

Solution Approach 1:

The patent segments the image analysis process into multiple stages: first extracting candidate regions of interest based on texture energy maps, then applying candidate masks to identify specific body parts, and finally using machine-learning models for verification. This segmentation allows the system to maintain speed by processing only relevant regions while achieving precise detection of specific body parts through hierarchical analysis.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent performs preliminary actions by extracting candidate regions of interest and generating texture energy maps before final body part detection. This preliminary processing filters out irrelevant areas and prepares data structures that accelerate subsequent verification steps, resolving the contradiction between speed and precision.

Inventive Principle:

Principle #10Preliminary action

2Measurement precision

If candidate masks and machine-learning models are applied for body part detection, then detection precision is improved, but device complexity increases

Engineering Contradiction:

Improvedetection precision of specific body partsVSAvoidsystem complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent divides the detection system into modular components: texture energy analysis unit, candidate region extraction unit, candidate mask application unit, and body part detection unit. Each module performs a specific function and can be independently optimized or replaced, reducing overall system complexity while maintaining high detection precision through specialized processing at each stage.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces candidate masks as intermediary elements between the texture energy maps and machine-learning models. These masks serve as mediators that translate abstract texture energy data into specific body part candidates, simplifying the interaction between processing stages and reducing computational complexity while improving detection accuracy.

Inventive Principle:

Principle #24Intermediary (Mediator)

3Adaptability or versatility

If texture energy maps are analyzed with multiple candidate masks, then adaptability to different body parts is improved, but processing time increases

Engineering Contradiction:

Improveadaptability to detect different body partsVSAvoidprocessing time

Core Design Contradiction:

Adaptability or versatilityVSLoss of time

Solution Approach 1:

The patent applies partial action by selectively applying candidate masks only to candidate regions of interest rather than processing the entire image with all masks. This approach maintains adaptability to detect various body parts while significantly reducing processing time by focusing computational resources only on relevant regions.

Inventive Principle:

Principle #16Partial or excessive action

Solution Approach 2:

The patent performs preliminary extraction of candidate regions of interest based on texture energy maps before applying candidate masks. This preliminary action filters out irrelevant areas and prepares the data structure in advance, enabling fast and adaptive detection of different body parts without time-consuming full-image processing.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS8620091B2Method and apparatus for detecting specific external human body parts from texture energy maps of images by performing convolution

Publication Date: 2013.12.31 ELECTRONICS & TELECOMM RES INST

AI summary

An apparatus for detecting specific human body parts in an image includes: a texture energy analysis unit for analyzing energy distribution in the image and generating texture energy maps; a candidate region-of-interest extraction unit for extracting candidate regions-of-interest for the specific body parts on a given texture energy map by applying a threshold to the given texture energy map, the given texture energy map being selected among the texture energy maps; a candidate mask application unit for performing convolution between candidate masks for the specific body parts and the candidate regions-of-interest and selecting candidate body parts based on results of the convolution; and a body part detection unit for detecting the specific body parts on the image by performing verification on the candidate body parts. The verification is performed by using machine-learning models for the specific body parts.