Automated Main Subject Extraction Using Superpixel Segmentation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing methods for extracting the main subject from images are inefficient and require manual intervention, making them unsuitable for processing large volumes of online product images, where the background can affect search results and manual segmentation is time-consuming.

Innovation Solution

A method and system that automatically extract the main subject from images by determining the presence of a specified feature, using superpixel segmentation, and applying algorithms like AdaBoost and GrabCut to identify and isolate the main subject region, improving efficiency and accuracy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual intervention methods are used for extracting main subject, then extraction accuracy can be maintained, but processing efficiency deteriorates significantly

Engineering Contradiction:

Improveextraction accuracyVSAvoidprocessing efficiency

Core Design Contradiction:

Measurement precisionVSProductivity

Solution Approach 1:

The system performs automatic main subject extraction without requiring manual intervention. The computer executes algorithms that autonomously identify and extract the main subject from product images, making the system self-sufficient and eliminating the need for human operators to manually select or frame segmented regions.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The patent replaces manual mechanical operations with automated computational algorithms. Instead of human operators manually framing and selecting segmented regions, the system uses computer-based image processing algorithms including superpixel segmentation, AdaBoost classification, and GrabCut extraction to automatically identify and extract the main subject.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Productivity

If automated extraction methods are implemented, then processing efficiency is improved, but extraction accuracy may deteriorate

Engineering Contradiction:

Improveprocessing efficiencyVSAvoidextraction accuracy

Core Design Contradiction:

ProductivityVSMeasurement precision

Solution Approach 1:

The system divides the image processing task into multiple sequential stages: superpixel segmentation to partition the image into meaningful regions, AdaBoost classification to identify candidate regions containing the main subject, and GrabCut extraction to precisely delineate the main subject boundaries. This multi-stage segmentation approach maintains high accuracy while enabling automated processing.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system introduces intermediate processing steps between initial image input and final extraction output. Superpixel segmentation creates intermediate regions, AdaBoost generates intermediate candidate selections, and these intermediates guide the final GrabCut extraction. These intermediary steps ensure accurate results while maintaining automated efficiency.

Inventive Principle:

Principle #24Intermediary (Mediator)

3Loss of information

If background regions are included in product images, then complete product context is preserved, but search result accuracy deteriorates

Engineering Contradiction:

Improveproduct context preservationVSAvoidsearch result accuracy

Core Design Contradiction:

Loss of informationVSMeasurement precision

Solution Approach 1:

The system extracts and removes background regions from product images, isolating only the main subject. By taking out the background and keeping only the essential product content, the system preserves the critical product information needed for accurate search results while eliminating distracting or irrelevant background elements that could interfere with search accuracy.

Inventive Principle:

Principle #2Taking out (Extraction)

Data Source

PatentUS10497121B2Method and system for extracting a main subject of an image

Publication Date: 2019.12.03 ALIBABA GROUP HOLDING LTD

AI summary

A method includes determining whether a region to be recognized corresponding to a specified feature exists in an image to be processed. The method may also include determining a main subject region containing the main subject of the image to be processed in accordance with a preset feature parameter of the main subject region and the coordinates and size of the region to be recognized when the region to be recognized exists. The method may further include extracting an image of the main subject region as a foreground image for an extraction process of a foreground target, and taking the extracted image as the main subject of the image to be processed.