Handwritten Matrix Recognition Using Spatial Relationships

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing handwriting recognition systems struggle to accurately recognize complex structures like matrices without relying on specific designation elements or gestures, leading to increased processing time and complexity, especially when dealing with complex elements, misalignments, or empty cells.

Innovation Solution

A method that recognizes matrices and similar structures by analyzing the relative positions and geometrical relationships of handwritten elements, without requiring specific designation elements, using a combination of preprocessing, segmentation, recognition, and language experts to achieve accurate and efficient recognition.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If handwriting recognition systems use indicative elements like brackets or spatial alignment to recognize matrix structures, then structure recognition accuracy is improved, but the system cannot handle complex elements, misaligned elements, or empty cells, reducing adaptability

Engineering Contradiction:
Improvestructure recognition accuracyVSAvoidability to handle complex matrices
Core Design Contradiction:
Measurement precisionVSAdaptability or versatility

Solution Approach 1:

The system segments the matrix recognition task into multiple independent analysis dimensions: row identification based on vertical alignment, column identification based on horizontal alignment, element type classification, and structural pattern matching. This segmentation allows each dimension to be processed separately, enabling the system to handle complex elements, misalignments, and empty cells without compromising overall recognition accuracy.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system changes the recognition parameters from relying on specific indicative elements (brackets, perfect alignment) to using multiple flexible parameters including relative positioning, spacing patterns, element type classifications, and contextual relationships. This parameter transformation enables the system to adapt to various matrix formats and complexities while maintaining recognition precision.

Inventive Principle:
Principle #35Parameter changes

2Measurement precision

If the system performs absolute recognition of matrix structures first, then structure identification is improved, but content recognition is influenced negatively and processing complexity increases

Engineering Contradiction:
Improvematrix structure recognitionVSAvoidrecognition process complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system merges structure recognition and content recognition into a unified parallel processing framework. Instead of performing absolute structure recognition first and then content recognition sequentially, the system simultaneously analyzes structural patterns and element contents, sharing computational resources and intermediate results between the two tasks. This merging reduces overall processing complexity while maintaining both structure and content recognition accuracy.

Inventive Principle:
Principle #5Merging (Combining)

3Measurement precision

If handwriting recognition systems process structured content considering relative positioning, then structure recognition is improved, but processing time increases

Engineering Contradiction:
Improvestructured content recognition accuracyVSAvoidprocessing time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The system performs preliminary analysis of relative positioning relationships and spatial patterns during the input phase, before full recognition processing begins. By pre-processing and storing positional information, alignment relationships, and spatial patterns in an optimized format, the system reduces the computational burden during the main recognition phase, thereby decreasing overall processing time while maintaining high recognition accuracy for structured content.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP3405906B1System and method for recognizing multiple object structure
Publication Date: 2025.08.20 MYSCRIPT
  • EP3405906B1 patent drawingFigure 1
  • EP3405906B1 patent drawingFigure 2
  • EP3405906B1 patent drawingFigure 3

AI summary

A system, method, and computer program for recognizing an arrangement of multiple objects on computing devices. The computing device may include a processor and at least one non-transitory computer readable medium for recognizing handwriting input under control of the processor. The non-transitory computer readable medium may be configured to determine with the medium at least one geometrical relationship between a plurality of recognized elements "of the input and allocate with the medium the recognized elements having the at least one geometrical relationship therebetween to corresponding positions of an arrangement of the recognized elements. The positions of the arrangement may be cells of a two-dimensional structure of the recognized elements. The at least one geometrical relationship may be an overlap of the plurality of recognized elements in one or more directions of the arrangement.