Multi-resolution spatial feature extraction for handwriting recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional handwriting recognition techniques face challenges in accurately recognizing characters due to variability in human behavior, which results in artifacts such as slants, hooks, and wiggles, leading to the need for lower resolution bitmaps that may obscure important details, especially in complex characters like Chinese characters.
Innovation Solution
The extraction of spatial features at multiple resolutions, where pixel content values are determined for sets of pixels at various distances and directions from a pixel, allowing for a more detailed representation of character shapes and improving recognition accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the resolution of the device receiving the handwritten characters is lowered (coarser bitmap), then artifacts such as slants, hooks, and wiggles are smoothed out, but important details of the input character are obscured
Solution Approach 1:
The patent segments the spatial feature extraction process into multiple resolution levels. Instead of using a single coarse bitmap, the system extracts features at multiple resolutions (e.g., 32x32, 64x64, 128x128 bitmaps), allowing each resolution level to capture different aspects of character information. This segmentation enables the system to maintain both smoothing of artifacts and preservation of important details.
Solution Approach 2:
The patent adds a resolution dimension to the feature extraction process. By extracting spatial features across multiple resolution levels rather than a single fixed resolution, the system transforms the problem from two-dimensional (single bitmap size) to three-dimensional (multiple bitmap sizes), enabling comprehensive capture of character information at different scales.
2Ease of manufacture
If a fixed resolution bitmap is used throughout feature extraction, then the processing is simple and consistent, but spatial features cannot capture detailed information at varying scales
Solution Approach 1:
The patent introduces dynamic resolution selection into the feature extraction process. Instead of using a static fixed-resolution bitmap, the system dynamically extracts features at multiple resolutions and combines them. This allows the system to adapt to different character sizes and complexities, improving measurement precision while maintaining processing efficiency through systematic multi-resolution analysis.
Data Source
AI summary
A first technique of recognizing content is disclosed, including: determining a first value representative of a pixel content present at a first set of pixels associated with a first distance from a pixel under consideration; determining a second value representative of a pixel content present at a second set of pixels associated with a second distance from the pixel under consideration; and using the first and second values to compute one or more spatial features associated with the pixel under consideration for purposes of content recognition. A second technique of recognizing content is also disclosed, including: determining, for a pixel, a first value representative of a first feature associated with a set of pixels associated with a first direction from the pixel; and determining, for the pixel, a second value representative of a second feature associated with a set of pixels associated with a second direction from the pixel.