Hough-Based Document Localization for Fast On-Device Scanning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing document recognition systems on mobile devices face challenges due to resource constraints, such as limited processing power and memory, leading to inefficiencies in response time and accuracy.
Innovation Solution
A document localization method using a Hough-based approach that detects edges and lines in images, calculates contour and contrast scores for candidate quadrilaterals, and refines them using similarity transformations to enhance accuracy and reduce computational load on mobile devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a traditional document recognition system is implemented on a mobile device, then the system can perform document recognition, but the response time increases and processing accuracy decreases due to limited processing power and memory
Solution Approach 1:
The patent segments the document recognition process into distinct modules: edge detection, line detection, quadrilateral candidate generation, scoring, and selection. This modular approach allows each component to be optimized independently for mobile devices, improving both speed and accuracy without requiring the entire system to be redesigned.
Solution Approach 2:
The patent performs preliminary actions by pre-calculating and storing edge information, line parameters, and scoring criteria before the actual document recognition occurs. This preparation reduces the computational burden during real-time processing, thereby decreasing response time while maintaining recognition accuracy on resource-constrained mobile devices.
2Measurement precision
If more computational resources are allocated to document recognition on mobile devices, then processing accuracy improves, but device complexity and power consumption increase
Solution Approach 1:
The patent replaces complex mechanical computation with mathematical transformations and geometric calculations. By using analytical solutions for quadrilateral scoring and validation based on geometric properties and contrast measurements, the system achieves high precision without requiring sophisticated hardware or complex processing algorithms.
Solution Approach 2:
The patent changes the parameters used for document localization from exhaustive pixel-level analysis to edge-based and line-based features. By transforming the problem from analyzing all pixels to analyzing detected edges and lines, the system reduces computational complexity while maintaining or improving localization accuracy through targeted parameter measurement.
3Measurement precision
If exhaustive search methods are used to find all candidate quadrilaterals, then localization accuracy improves, but processing time and computational load increase significantly
Solution Approach 1:
The patent applies partial action by generating and evaluating only the necessary quadrilateral candidates rather than exhaustively searching all possible combinations. By using edge and line detection to guide candidate generation, the system processes only relevant portions of the image data, achieving accurate document localization without the computational burden of exhaustive search.
Solution Approach 2:
The patent uses copying by creating simplified representations of document candidates through quadrilateral approximations rather than processing the full image data for each potential document location. This copying approach allows rapid evaluation of multiple candidates through scoring functions that work on simplified geometric representations, significantly improving processing speed while maintaining detection accuracy.
Data Source
AI summary
Advanced Hough-Based On-Device Document Localization. In an embodiment, lines are detected in an input image of a document. The lines are searched for candidate quadrilaterals. For at least a subset of the found candidate quadrilaterals, a contour score is calculated, and the candidate quadrilaterals are saved or discarded based on their contour scores. For each saved candidate quadrilateral, a contrast score is calculated. A final candidate quadrilateral is selected, based on the combined contour and contrast scores for the saved candidate quadrilaterals, to represent the borders of the document.