Coupled Saliency-Map Object Extraction via Adaptive Tri-Maps
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current automated object extraction methods from images are either labor-intensive or lack accuracy, with user-guided methods requiring significant manpower and fully automated methods failing to provide precise object extraction without advanced information.
Innovation Solution
An apparatus and method utilizing a coupled saliency-map, adaptive tri-map, and alpha matte generation, where a coupled saliency-map is generated by combining global and local saliency-maps, and an adaptive tri-map is created through Gaussian blur and image clustering, enabling automatic object extraction with user-correction capabilities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If user-guided matting method is used to extract salient objects, then extraction precision is improved, but manual labor cost increases significantly
Solution Approach 1:
The patent segments the object extraction process into three distinct modules: tri-map generation (automated), alpha matte generation (automated computation), and refinement (user-guided only where needed). This segmentation allows the majority of processing to be automated while preserving user control for critical decisions, thereby reducing manual labor while maintaining high extraction precision.
Solution Approach 2:
The system performs preliminary automated actions by generating the tri-map and initial alpha matte before user intervention is required. This preliminary automated processing handles the computationally intensive and routine aspects of extraction, reducing the subsequent manual work needed while preserving overall precision.
2Loss of time
If fully automated object extraction method is used, then manual labor cost is reduced, but extraction precision deteriorates without user intervention
Solution Approach 1:
The patent implements a feedback mechanism where user corrections to the tri-map or alpha matte are fed back into the system to refine the extraction results. This feedback loop allows the automated system to learn from and adapt to user preferences, improving precision while maintaining automation. The system can automatically adjust parameters based on user feedback without requiring complete manual re-processing.
3Measurement precision
If tri-map is generated manually for each image to improve extraction accuracy, then object extraction precision is improved, but processing time and labor cost increase
Solution Approach 1:
The patent replaces the manual mechanical process of tri-map generation with an automated computational algorithm. The system uses image processing techniques including Gaussian blur, edge detection, and region segmentation to automatically generate accurate tri-maps, substituting human manual work with automated mechanical/computational processes that are both fast and precise.
Solution Approach 2:
The system automatically adjusts key parameters such as Gaussian blur sigma values, threshold levels, and region segmentation parameters based on image characteristics. This dynamic parameter adjustment allows the automated tri-map generation to adapt to different images while maintaining high precision, eliminating the need for manual parameter tuning for each image.
Data Source
AI summary
According to one general aspect, an apparatus for extracting an object includes an image receiver configured to receive an image; a coupled saliency-map generator configured to generate a coupled saliency-map which is the sum of the product of a global saliency-map of the image and a predetermined weight value and a local saliency-map; an adaptive tri-map generator configured to generate an adaptive tri-map corresponding to the coupled saliency-map; an alpha matte generator configured to generate an alpha matte based on the adaptive tri-map; and an object detector configured to extract an object according to transparency of the alpha matte to generate an object image.