Image Feature Detection via Thumbnail Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for identifying interesting features in computer content are cumbersome and raise privacy concerns, as they often require processing large amounts of data or sending entire images to remote servers, which can expose personal information.
Innovation Solution
A method that reduces pixel and color resolution of captured images, processes them locally to isolate and identify interesting features, and extracts a sub-image for analysis, thereby reducing data processing demands and maintaining user privacy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the entire displayed content is processed to identify interesting features, then feature detection accuracy is improved, but user privacy is compromised and data processing demands increase
Solution Approach 1:
The patent divides the image processing task into two stages: first, a low-resolution thumbnail is processed to identify the region of interest; second, only that specific region is extracted from the high-resolution original image for detailed analysis. This segmentation allows accurate feature detection while minimizing the amount of sensitive data processed or transmitted.
Solution Approach 2:
The patent extracts only the relevant sub-image containing the interesting feature from the full high-resolution image, rather than processing or transmitting the entire image. This extraction principle reduces data processing demands and protects user privacy by isolating and transmitting only the necessary portion.
2Measurement precision
If high-resolution images are processed to identify interesting features, then detection accuracy is improved, but data processing demands and computational resources increase
Solution Approach 1:
The patent segments the processing workload by using a low-resolution thumbnail for initial feature detection and region identification. Once the region of interest is identified in the thumbnail, only that specific region is extracted from the high-resolution original image, significantly reducing the amount of data that requires intensive processing.
Solution Approach 2:
The patent extracts only the necessary sub-image region from the full high-resolution image after identifying the region of interest in the thumbnail. This extraction approach minimizes computational resources and data processing demands while maintaining detection accuracy for the interesting feature.
3Measurement precision
If manual image capture and processing is used to find related images, then feature identification accuracy is improved, but user effort and time consumption increase
Solution Approach 1:
The patent implements an automated system that performs the complete workflow of capturing the display content, generating a thumbnail, identifying interesting features, extracting the relevant sub-image, and submitting it for related image search. This self-service automation eliminates manual user effort while maintaining accurate feature identification through the thumbnail-based detection approach.
Data Source
AI summary
Systems and methods for identifying an interesting feature in displayed content are presented. In response to a user's direction to identify an interesting feature of content displayed on the user's computing device, an image capture operation is conducted. The image capture operation captures some of the content displayed on the computing device and rendered to a working image. One or more processing passes are conducted on the working image to reduce various aspects of the working image. Remaining features in the working image, after the one or more processing passes, are identified and at least one feature is identified as a selected feature. A bounding box of the selected feature is mapped to the captured image, and a sub-image of the captured image as determined by the mapped bounding box is used as an interesting feature of the captured image.