Targeted Image Detection Across Computing Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for detecting specific images within computing networks, such as the Internet and virtual reality environments like the Metaverse, are inadequate due to the dynamic nature of displayed images and the challenge of navigating static network environments, especially when images change over time or are altered to evade detection.
Innovation Solution
A comprehensive system that navigates to predetermined websites to access HTML code, searches for image and video file extensions, and employs a virtual surveillance drone within virtual reality environments to capture and analyze images for matches with targeted images, using machine learning to identify patterns of variance and unauthorized use.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If screenshot capture method is used to detect images on websites, then detection simplicity is improved, but detection completeness deteriorates due to rotating banners and dynamic image changes
Solution Approach 1:
The patent transitions from static screenshot capture to dynamic webpage navigation and rendering. The system actively navigates to webpages, triggers image loading through browser rendering, and captures images during the dynamic display process, ensuring detection of rotating banners and dynamically changing images that static screenshots would miss.
Solution Approach 2:
The system implements continuous monitoring by repeatedly navigating to webpages and capturing images during their display cycles. This continuous action ensures that rotating banners and dynamically updated images are detected across multiple display cycles, maintaining detection completeness while automating the process.
2Productivity
If periodic screenshot capture is performed to detect changing images, then detection frequency is improved, but time consumption and resource usage worsen
Solution Approach 1:
The system performs preliminary actions by navigating to webpages and triggering image loading before actual detection occurs. By pre-loading images through browser navigation and rendering, the system captures images during their natural display cycle rather than relying on periodic external screenshot captures, improving detection frequency while reducing redundant time consumption.
3Reliability
If image detection system searches all websites comprehensively, then detection coverage is improved, but system complexity and computational resources worsen
Solution Approach 1:
The patent segments the image detection task into distinct modular components: webpage navigation module, image capture module, image processing module, and result analysis module. Each module handles a specific aspect of the detection process, reducing overall system complexity while maintaining comprehensive detection coverage through coordinated operation of these segmented functions.
Solution Approach 2:
The system introduces a browser environment as an intermediary between the detection system and target websites. The browser acts as a mediator that handles webpage rendering, image loading, and display, allowing the detection system to leverage existing browser capabilities rather than building complex webpage interaction logic from scratch, thus reducing system complexity while maintaining detection coverage.
4Measurement precision
If traditional image matching is used to detect exact matches only, then matching precision is improved, but adaptability to image variations deteriorates
Solution Approach 1:
The patent applies parameter changes by transforming images into frequency domain representations (e.g., Fourier transforms) and analyzing spectral characteristics rather than comparing pixel values directly. This parameter transformation enables the system to maintain high matching precision for exact matches while becoming adaptable to variations in scale, rotation, and other transformations that preserve spectral properties.
Data Source
AI summary
Detecting targeted images, as well as variations thereof within computing networks, such as the Internet and the Metaverse. Navigation to predetermined websites and accessing the HTML code to search for image file extensions and, in some instances video file extensions within the listing of URLs. In response to finding URLs with image or video file extensions, the hyperlink is activated to navigate to the webpage containing the associated image or video and the image or video is downloaded for purposes of target image detection analysis. A virtual surveillance drone is deployed within virtual reality computing systems to navigate the entirety of the environment to search for images or, in some instances videos that may be displayed and, more specifically, images that appear to match or resemble the targeted image. The drone is capable of capturing the images and communicating the images to non-virtual computing systems for purposes of target image detection analysis.