Camera Tracking for OCR Using Feature Point Comparison and Trace
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing mobile terminal systems face challenges in accurately tracking and processing large paper documents using camera-based OCR, as the resolution of captured images is often insufficient, leading to positional deviations and errors in feature point tracing, which can result in incomplete or inaccurate OCR results.
Innovation Solution
The system employs a combination of feature point comparison processing and feature point trace processing to identify and track the OCR processing target area, using transformation matrices to map the data input area across images, while suppressing unnecessary feature point comparisons when camera movement exceeds a threshold, thereby minimizing tracking errors and maintaining image capturing rate.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If feature point comparison processing is performed continuously to track OCR target area, then tracking accuracy is improved, but processing time increases and image capturing rate decreases
Solution Approach 1:
The system performs feature point comparison processing periodically rather than continuously. A suppression determination unit decides whether to suppress comparison processing based on whether the captured image differs significantly from the reference image. When suppression is determined, the system skips comparison processing and uses only trace processing, thereby reducing processing time while maintaining adequate tracking accuracy.
2Productivity
If feature point trace processing is used to maintain high capturing rate, then processing speed is improved, but positional deviation accumulates over time
Solution Approach 1:
The system uses a feedback mechanism where the suppression determination unit continuously monitors the difference between captured images and reference images. When the difference exceeds a threshold, the system activates feature point comparison processing to correct accumulated positional deviations, then uses trace processing for intermediate frames. This feedback loop prevents error accumulation while maintaining high capturing rates.
3Measurement precision
If feature point comparison processing is performed frequently to correct positional deviation, then position accuracy is improved, but processing load increases and capturing rate decreases
Solution Approach 1:
The system applies partial comparison processing only when necessary. Instead of performing full comparison processing on every frame, the suppression determination unit selectively activates comparison processing only when image differences indicate significant positional changes. This partial action approach maintains position accuracy while minimizing processing load.
4Reliability
If the system adapts to large camera movements by frequent comparison processing, then tracking reliability is improved, but processing time increases
Solution Approach 1:
The system dynamically adjusts its processing strategy based on real-time conditions. The suppression determination unit evaluates image differences and dynamically switches between comparison processing and trace processing modes. This dynamic adaptation maintains tracking reliability during large camera movements while minimizing processing time during stable conditions.
Data Source
AI summary
First coordinate transformation information between an entire image and a first captured image is calculated by a feature point comparing process. Second coordinate transformation information between the first captured image and a second captured image is calculated by a feature point tracing process, the second captured image being a captured image at a timing when the first coordinate transformation information is calculated. Third coordinate transformation information between an immediately previous captured image and a third captured image is calculated by a feature point tracing process. A data input area in the entire image is mapped on the third captured image based on the first to the third coordinate transformation information pieces. Updates of the first and the second coordinate transformation information pieces may be suppressed where a change amount exceeds a predetermined threshold.