Image-Based Web Search System Using Feature Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional Internet search methods require users to know specific URL information or keywords, leading to inefficiencies when users forget URLs or lack correct search terms, resulting in significant time spent searching for desired information without alternative methods.
Innovation Solution
A search system utilizing images, where users can upload images to an image search server, which converts and compares image features like unit colors, patterns, and hash values to identify matching website information, allowing for image-based searching with adjustable error tolerance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If users use conventional search methods (URL or keywords), then they can access websites, but they spend excessive time searching when they forget URLs or lack correct keywords
Solution Approach 1:
The patent replaces the mechanical text-based search system (typing URLs or keywords) with an image-based search system. Users can upload images to search for websites, eliminating the need to manually type or remember text strings. The image conversion section transforms uploaded images into search format images, and the image search section compares these with stored images to retrieve relevant website information, thereby substituting the traditional text-input mechanism with an image-input mechanism that is more intuitive and time-efficient.
Solution Approach 2:
The patent uses image copying and comparison to enable search functionality. Instead of requiring users to create or recall text descriptions, the system accepts image copies/uploads and compares them against a database of stored images. The image search section detects matching images by comparing features such as color histograms, patterns, and structural characteristics, allowing users to search websites by uploading a copy or screenshot of the desired site rather than typing its URL or keywords.
2Productivity
If users upload images for search, then search time is reduced, but the system complexity increases due to image conversion and comparison processes
Solution Approach 1:
The patent segments the image search system into distinct functional modules to manage complexity. The image conversion section is divided into sub-modules for format conversion, resizing, and feature extraction. The image search section is segmented into comparison algorithms that analyze specific image characteristics separately (color histograms, patterns, structural features). This modular segmentation allows each component to handle a specific task efficiently, making the overall complex system more manageable and maintainable while preserving high search efficiency.
Solution Approach 2:
The patent transforms images into different parameter representations to facilitate efficient comparison. The image conversion section converts images into standardized formats with controlled dimensions and extracted feature parameters (such as color histograms, pattern descriptors, and structural metrics). By changing the representation from raw pixel data to standardized parameter sets, the system reduces the computational complexity of comparison operations while maintaining search accuracy and efficiency.
Data Source
AI summary
A search system using images is provided in which when a user does not know a relevant URL or search keyword correctly while surfing the Internet, he or she can search a desired website using only an image. The search system using images according to the present invention comprises an image search server and a user terminal. The image search system comprises: an image conversion section for converting the image included in the website information and the to-be-searched image uploaded by the user into search format images; an image search section for comparing eigen values of both the to-be-searched image uploaded by the user and the search image included in the website information and detecting the website information having a matching eigen value; and a storage section for storing the detected website information, the image included in the website information, and information regarding eigen values.