OCR Data Caching for Image Retrieval Speed
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional user devices with OCR functionality face inefficiencies in recognizing and retrieving text data from image files, requiring repetitive OCR processes that increase waiting time and inconvenience users.
Innovation Solution
A method and apparatus that store OCR data for image files, allowing quick retrieval and display of desired objects by matching stored OCR data with image files, reducing the need for repeated OCR processing and minimizing waiting time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If OCR function is executed every time to obtain objects from image files, then recognition accuracy is maintained, but recognition time and user waiting time increase significantly
Solution Approach 1:
The patent performs OCR recognition in advance and stores the extracted text data along with image file information before the user actually needs it. This preliminary action creates a cache of pre-processed data that can be quickly retrieved later, eliminating the need to repeat the time-consuming OCR process while maintaining recognition accuracy.
Solution Approach 2:
The patent selectively applies OCR processing only to specific image files that contain text objects of interest, rather than processing all image files uniformly. By identifying and processing only the relevant subset of images, the system maintains high recognition accuracy for needed data while minimizing unnecessary processing time and resource consumption.
2Reliability
If OCR process is repeated for the same image files, then up-to-date OCR data is obtained, but user convenience deteriorates due to repeated waiting time
Solution Approach 1:
The system pre-processes and stores OCR data from image files in advance, creating a cached repository of text information. When users later search for or need objects from these images, the system retrieves the pre-extracted text data instead of re-running OCR, thereby maintaining data reliability while significantly improving user convenience by eliminating repeated waiting periods.
Solution Approach 2:
The patent creates copies of the extracted text data and stores them separately from the original image files. These text copies can be quickly searched and retrieved without needing to access or re-process the original images, providing users with fast response times while ensuring the copied text data reflects accurate OCR results.
3Speed
If OCR data is stored for multiple image files, then quick retrieval is enabled, but device storage requirements increase
Solution Approach 1:
The patent extracts only the essential text information from image files using OCR and stores this extracted text data separately, rather than storing entire image files. This extraction approach significantly reduces the storage space required while enabling fast text-based search and retrieval operations, as text data occupies minimal space compared to full-resolution images.
Solution Approach 2:
The system segments the data storage into two distinct parts: the original image files are stored in their primary location, while separate text data extracted via OCR is stored in a dedicated cache or database. This segmentation allows independent management of storage resources, enabling quick text retrieval without requiring access to large image files, thus improving speed while controlling storage usage.
Data Source
AI summary
A method for processing data by using an optical character reader (OCR) is provided. The method includes obtaining OCR data from each image file of a plurality of image files and storing the obtained OCR data, receiving a search command with respect to an object, extracting the object from the stored OCR data, selecting OCR data which includes the object from among the OCR data, and displaying a list of image files which correspond to the selected OCR data.