Image Retrieval Using Extended Difference Degrees
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image retrieval systems face inefficiencies and inaccuracies due to information loss when determining difference degrees between images, which affects the overall efficiency and accuracy of the retrieval process.
Innovation Solution
The implementation of a neural network with a hash coding layer and a binary coding layer to determine extended subsets and extended difference degrees between images, improving processing speed and storage efficiency, and enhancing the accuracy of image retrieval by identifying result images based on these extended differences.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional difference degree analysis is used to identify result images, then the retrieval process is simpler, but information loss occurs which reduces accuracy
Solution Approach 1:
The patent extends the traditional difference degree analysis by introducing extended subsets that incorporate multiple reference images. Instead of comparing candidate images to a single target image, the system creates extended subsets containing target images and their related reference images, then calculates extended difference degrees based on these multi-dimensional relationships. This dimensional expansion preserves more information and improves retrieval accuracy.
Solution Approach 2:
The patent performs preliminary actions by pre-establishing extended subsets and calculating extended difference degrees before final image retrieval. The system pre-processes reference images to create extended subsets, and pre-calculates the extended difference degrees between target images and reference images. This preliminary processing preserves information that would otherwise be lost during the actual retrieval operation.
2Measurement precision
If extended subsets and extended difference degrees are calculated, then retrieval accuracy improves, but computational complexity increases
Solution Approach 1:
The patent segments the complex retrieval process into distinct modules: (1) acquiring reference images and organizing them into sets, (2) determining extended subsets by combining target images with reference images, (3) calculating extended difference degrees based on these subsets, and (4) identifying result images using the extended degrees. This segmentation makes the complex process more manageable and enables optimized processing at each stage.
Solution Approach 2:
The patent applies partial action by selectively extending the difference degree calculation only for images that require more precise differentiation. Instead of calculating extended difference degrees for all candidate images uniformly, the system applies the extended methodology selectively to cases where traditional analysis shows ambiguity or where high precision is critical, thus reducing overall computational burden while maintaining accuracy where needed.
3Loss of information
If extended difference degrees are used for image identification, then information loss is reduced, but processing time increases
Solution Approach 1:
The patent performs preliminary calculations of extended difference degrees between target images and reference images before the actual retrieval operation. By pre-processing and storing these extended difference degrees, the system avoids recalculating them during the retrieval phase, thus reducing real-time processing time while preserving the information benefits of extended analysis.
Solution Approach 2:
The patent creates copies of target images combined with reference images to form extended subsets. Instead of performing complex calculations on original images during retrieval, the system uses pre-generated extended subset copies that encapsulate the necessary information, enabling faster comparison and identification while maintaining the information integrity of the extended analysis.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present disclosure relates to systems and methods for image retrieval. The system may determine a plurality of difference degrees associated with an image set including a plurality of candidate images and a target image, each of the plurality of difference degrees corresponding to two images in the image set. For each image in the image set, the system may determine an extended subset based on the plurality of difference degrees. For each of the plurality of candidate images, the system may determine an extended difference degree between the candidate image and the target image based on an extended subset corresponding to the candidate image and an extended subset corresponding to the target image. The system may identify a result image corresponding to the target image from the plurality of candidate images based on the extended difference degrees corresponding to the plurality of candidate images.