Image Search via Text Labels and Server Resource Utilization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image searching technologies based on visual features do not effectively utilize server resources, as they require dedicated visual feature extraction and indexing, limiting the functionality of text searching engines.
Innovation Solution
Generating labels for images based on description information and storing corresponding relationships, allowing text searching engines to perform image searches, thereby improving server resource utilization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If visual feature based image searching database is established, then image searching function is achieved, but server resource utilization is poor
Solution Approach 1:
The patent applies universality by enabling text searching engines to perform image searching functions through label generation. The text searching engine is enhanced to handle both text and image search queries by converting image description information into searchable labels, allowing a single engine to serve multiple search purposes without requiring separate dedicated visual feature extraction systems.
Solution Approach 2:
The patent introduces label generation as an intermediary mechanism between image description information and text searching. By converting image metadata into standardized labels that can be indexed and searched by text engines, the system creates a bridge that allows text searching infrastructure to handle image search requests efficiently.
2Measurement precision
If dedicated visual feature extraction system is established, then image searching accuracy is improved, but device complexity increases
Solution Approach 1:
The patent extracts the essential search functionality from complex visual feature extraction systems and implements it through simpler label-based indexing. By separating the image storage/description from the search indexing process, the system maintains search accuracy while reducing the complexity of the search engine itself, as it only needs to process text labels rather than complex visual features.
Solution Approach 2:
The patent uses simple text labels as disposable intermediaries to represent image content for searching purposes. These labels are generated from image description information and used solely for indexing and search matching, then discarded after use. This approach avoids the need for complex, persistent visual feature extraction and reconstruction systems.
Data Source
AI summary
The present disclosure introduces a method and an apparatus for searching images. With respect to each image in an image searching database, respective labels of respective images are generated based on description information corresponding to the respective images. A corresponding relationship between the generated respective labels and the respective images is stored. Based on a received image searching request, description information corresponding to an image for search in the image searching request is obtained. Based on the description information of the image for search, the label of the image for search is generated. Based on the stored corresponding relationship between the respective labels and the respective images, one or more images corresponding to the label of the image for search are determined. The determined one or more images are sent to the client terminal that sends the image searching request.