Fragmented Image Search Matching for Heterogeneous Databases
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing search matching methods for large-scale heterogeneous databases in cloud-based object recognition services are inefficient and resource-intensive, leading to significant operating times and potential misidentification errors due to the complexity and diversity of data arrays across multiple service providers and installations.
Innovation Solution
A high-speed search matching method utilizing a fragmented database structure, initialized with matrix and data-table objects, employs block-based computing and a marker matrix to optimize performance and accuracy by processing image features and accompanying information efficiently, ensuring precise searches within defined search groups.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a traditional search matching method is used for large-scale heterogeneous databases, then the system can handle multiple service providers and users, but the search performance and processing speed deteriorate due to the need to calculate distances across the entire database
Solution Approach 1:
The patent divides the large-scale heterogeneous database into multiple independent data arrays, each corresponding to a specific service provider or user. The search engine determines the target data array based on the image sample's service provider identifier, allowing searches to be confined to relevant subsets rather than the entire database. This segmentation resolves the contradiction by maintaining adaptability across multiple providers while improving productivity through reduced search scope.
Solution Approach 2:
The patent implements local quality by creating specialized data arrays with specific characteristics for each service provider or user. Each data array is optimized for its specific context, and the search engine selects the appropriate array based on the query requirements. This allows the system to maintain versatility for different providers while achieving high search performance within each localized data context.
2Reliability
If the search engine searches the entire database for each image sample, then all possible matches are found, but the calculation time and resource consumption increase significantly
Solution Approach 1:
The patent extracts and identifies the relevant target data array from the large-scale database based on the image sample's attributes (such as service provider identifier). By taking out only the necessary subset of data for the current search task, the system maintains reliability by ensuring all relevant matches are found within the target array, while significantly reducing calculation time by avoiding unnecessary computations on unrelated data.
Solution Approach 2:
The patent performs preliminary actions by pre-organizing the database into structured data arrays and pre-determining the target array selection criteria. Before actual search execution, the system identifies which data array should be searched based on the image sample's characteristics. This preliminary action ensures completeness of results while minimizing computation time during the actual search phase.
3Ease of manufacture
If a cyclic calculation method is used for distance computation, then the implementation is simple and compatible with edge devices, but the processing speed is insufficient for large-scale databases
Solution Approach 1:
The patent applies segmentation by dividing the large-scale search task into multiple smaller searches across different data arrays. This allows the use of simple cyclic calculation methods within each smaller array, maintaining implementation simplicity and device compatibility, while the overall processing speed improves because the total computation is distributed across multiple focused searches rather than one large exhaustive search.
Data Source
AI summary
The high-speed search matching method with a large-scale heterogeneous database includes the process of extracting and storing the user's registered image features, forming a database, and being a reference when matching and searching; The process of initializing the search engine, in which two objects in the search engine are initialized, including a matching object, calculating the distance, and a data table object to determine the target search object for each input image sample; The process of matching and searching for objects, including the process of calculating the distance between the features and determining the target array, and finally comparing and returning the final result. The proposed method has demonstrated the ability to calculate accurately as well as very high performance, especially in large-scale data processing cases.