Document Image Capture via Block Segmentation and Parameter Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current optical image capture systems face challenges in capturing high-quality images of large documents, leading to inefficient data extraction due to suboptimal focus and exposure settings across the document, requiring multiple image captures and increased bandwidth and processing power, while also posing security risks.
Innovation Solution
A system and method that divides the document into image blocks, optimizes image capture parameters like focus and exposure for each block, captures and aggregates these images to reconstruct a high-quality document image, and transfers it efficiently to a data extraction engine.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If the camera is positioned a significant distance away from the source document to capture the entire document, then the entire document can be captured in the image, but the image capture parameters (focus and exposure) are only optimized for a portion of the document, resulting in poor quality images of other portions
Solution Approach 1:
The patent divides the document into multiple image blocks (e.g., quadrants or regions) and captures each block separately with optimized camera parameters for that specific region. This segmentation allows the camera to maintain high focus and exposure settings for each individual block while still covering the entire document across multiple captures.
Solution Approach 2:
The patent applies different image capture parameters (focus, exposure, white balance) to different regions of the document based on local lighting conditions and document characteristics. Each image block is captured with parameters optimized for its specific location, ensuring high image quality across the entire document surface.
2Reliability
If multiple images of the source document are captured to ensure all portions are clearly captured, then complete document data can be obtained, but the bandwidth required to transfer multiple images and the processor cycles required to process them increase significantly
Solution Approach 1:
The patent extracts only the necessary image blocks containing relevant data from the document for processing and transfer. Instead of transferring complete multiple images, the system identifies and extracts specific regions of interest (image blocks) that contain the required information, reducing the total data volume that needs to be transmitted and processed.
Solution Approach 2:
The patent performs preliminary processing of captured images to identify and extract only the necessary image blocks containing relevant data before transfer. This preliminary action reduces the amount of data that needs to be transmitted and processed by the data extraction engine, conserving bandwidth and processing power.
3Reliability
If multiple images of the source document are transferred to a remote data extraction engine, then complete document data can be processed, but security risks increase particularly for sensitive documents
Solution Approach 1:
The patent extracts and processes only the necessary image blocks locally or with minimal transfer to remote systems. By extracting only the essential data-containing regions rather than transferring complete multiple images to remote servers, the system reduces the security exposure for sensitive documents while maintaining data extraction completeness.
4Reliability
If the user is forced to capture multiple images of the source document to obtain complete data, then all document information can be captured, but the process becomes frustrating and time-consuming for the user
Solution Approach 1:
The patent implements automatic image block capture and processing without requiring manual user intervention. The system automatically divides the document into image blocks, captures each block with optimized parameters, and processes them for data extraction. This self-service approach eliminates the need for users to manually capture multiple images, significantly reducing user effort and frustration.
Solution Approach 2:
The patent performs preliminary automatic processing to determine which image blocks contain relevant data and captures only those necessary blocks. This preliminary action reduces the number of images the user needs to capture and provides a more streamlined, user-friendly experience while ensuring complete data extraction.
Data Source
Figure 1
Figure 2A
Figure 2B
AI summary
A viewfinder screen display is generated and positioned such that a source document is displayed in the viewfinder screen display. Source document image blocks corresponding to different portions of the source document are then defined. For each source document image block, the image capture parameter of an image capture device is set to an optimized image captare parameter setting for the source document image block. The image capture device then captures an image block optimized image of the source document optimized for the source document image block. The optimized source document image blocks are then extracted from each image block optimized image of the source document. The extracted optimized source document image blocks are then aggregated and used to construct an image capture parameter optimized, image of the source document.