Target Image Extraction Using Eye-Tracking and Mouse Coordinates
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image processing methods fail to accurately determine the target image desired by the user, and there is a lack of technologies to assist in finding the target image using user operation data like eye-tracking or mouse clicks.
Innovation Solution
A method involving eye-tracking and mouse operation data to obtain a background screenshot, extract candidate images, apply edge extraction and image entropy analysis to identify the target image, and set a target coordinate based on the image centroid.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If image analysis and data analysis algorithms are used to extract images from pictures or videos, then image extraction capability is improved, but the ability to accurately determine the target image desired by the user deteriorates
Solution Approach 1:
The patent introduces user operation behavior data (eye-tracking coordinates, mouse click coordinates) as an intermediary element that bridges the gap between automatic image extraction algorithms and user intent. This intermediary data serves as a mediator that guides the system to identify the specific target image the user wants, resolving the contradiction between automated extraction capability and accurate target determination.
Solution Approach 2:
The patent implements a feedback mechanism where user operation behavior data is continuously collected and used to refine and adjust the target image determination process. By feeding back eye-tracking and mouse click information to the image extraction system, the algorithm can iteratively improve its accuracy in identifying the user's desired target image, thus resolving the precision issue while maintaining high extraction capability.
2Loss of information
If clear user operation instructions such as mouse click coordinate or eye-tracking focus coordinate are provided, then user intent is clarified, but technology to assist in finding the target image is lacking
Solution Approach 1:
The patent enables the system to automatically process and utilize user operation behavior data without requiring additional user intervention. The system self-services by automatically integrating eye-tracking and mouse click coordinates with image extraction algorithms to locate and identify the target image, eliminating the need for separate assistance tools while maintaining ease of operation.
Solution Approach 2:
The patent merges multiple data sources including user operation behavior data (eye-tracking, mouse clicks), image extraction results, and coordinate information into a unified processing framework. By combining these previously separate elements into an integrated system, the patent achieves both clear user intent recognition and automated target image location assistance without requiring additional operational complexity.
3Measurement precision
If multiple processing steps including edge extraction and image entropy analysis are applied, then target image identification accuracy is improved, but processing complexity increases
Solution Approach 1:
The patent applies preliminary actions by pre-processing images with edge extraction and calculating image entropy before final target identification. These preparatory steps organize and enhance the image data in advance, making the subsequent target image identification more accurate while structuring the processing flow to manage complexity systematically.
Solution Approach 2:
The patent segments the image processing task into distinct functional modules: edge extraction module, image entropy analysis module, coordinate processing module, and target identification module. By dividing the complex processing into separate manageable segments, each performing a specific function, the system achieves high identification accuracy while the modular structure helps manage and reduce overall processing complexity.
Data Source
AI summary
A method, an apparatus and a computer-readable storage medium for target image extraction. The method comprises obtaining specified coordinate based on user operation behavior data, capturing a background screenshot based on the specified coordinate, and finally determining the target image that the user wants from the background screenshot.