Image Search Using Semantic Sub-section Segmentation and ANN Layer Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image search methods are inefficient and limited in scope, as they require processing and searching the entire input image, which consumes more memory and processing power, and do not support searching specific regions of interest or providing user-desired results, especially for small regions.
Innovation Solution
The method involves identifying sub-sections of an input image based on semantic parameters, determining optimal resolution, and selecting an optimal set of layers for an Artificial Neural Network (ANN) model to perform image search, allowing for efficient and pattern-based image searching by processing only relevant parts of the image.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional approaches process the complete input image for searching, then comprehensive search coverage is achieved, but processing time and memory consumption increase significantly
Solution Approach 1:
The patent divides the complete input image into multiple sub-sections or regions of interest. Instead of processing the entire image, the system identifies and processes only relevant sub-sections that contain objects matching user search criteria. This segmentation approach maintains comprehensive search coverage for relevant areas while significantly reducing processing time and computational resources for the complete image.
Solution Approach 2:
The patent extracts and processes only the necessary portions of the image that contain relevant search objects. By identifying regions of interest and extracting those specific sub-sections for processing, the system achieves comprehensive search coverage for target objects without the computational overhead of processing the entire image, thus reducing processing time while maintaining search effectiveness.
2Quantity of substance
If conventional approaches compress the complete input image for storing and searching, then storage efficiency is improved, but search precision and ability to search specific regions deteriorate
Solution Approach 1:
The patent segments the complete input image into multiple sub-sections and processes or stores only the relevant sub-sections containing objects of interest. This approach improves storage efficiency by avoiding compression and storage of entire images when only specific regions are needed for search, while maintaining search precision by preserving the quality and detail of relevant sub-sections.
Solution Approach 2:
The patent extracts relevant sub-sections from the complete input image for storage and searching. By taking out only the necessary portions containing search-target objects, the system achieves better storage efficiency without compromising search precision, as the extracted sub-sections maintain their original quality and can be searched with high accuracy.
3Reliability
If conventional approaches process the entire input image, then all potential matches are found, but processing power and memory requirements increase
Solution Approach 1:
The patent divides the input image into sub-sections and identifies regions containing objects that match user search criteria. By processing only these relevant sub-sections instead of the entire image, the system maintains match completeness for target objects while significantly reducing processing power and memory requirements, as computational resources are focused only on relevant image regions.
Solution Approach 2:
The patent applies different processing quality and attention to different regions of the image based on their relevance to the search query. Regions containing objects of interest receive full processing attention to ensure match completeness, while irrelevant regions are either skipped or processed at lower quality, thereby reducing overall processing power requirements while maintaining reliability for relevant matches.
4Productivity
If conventional approaches search the complete image, then comprehensive results are provided, but processing time and energy consumption increase
Solution Approach 1:
The patent segments the complete image into sub-sections and processes only those containing objects relevant to user search criteria. This maintains search comprehensiveness for target objects while reducing energy consumption by avoiding processing of irrelevant image regions, as the system focuses computational energy only on meaningful search areas.
Solution Approach 2:
The patent extracts and processes only the necessary sub-sections containing search-target objects from the complete image. This approach provides comprehensive search results for relevant objects while minimizing energy consumption by excluding unnecessary processing of irrelevant image portions, thereby improving productivity per unit of energy consumed.
Data Source
AI summary
The disclosure relates to method and system for performing optimized image search. The method includes receiving an input image and user requirements with respect to an image search, identifying a sub-section from various sub-sections of the input image based on semantic parameters corresponding to each sub-section and the user requirements, and determining an optimal resolution of the sub-section based on a pixel density of various image formats derived for the sub-section. The method further includes identifying an optimal set of layers from a plurality of layers of an Artificial Neural Network (ANN) based image search model for performing the image search based on the semantic parameters for the sub-section, the optimal resolution of the sub-section, and historical data, and performing the image search to identify a set of output images similar to the sub-section based on a modified ANN based image search model comprising the optimal set of layers.