2D Code Recognition Using Block Segmentation and Weighting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for recognizing 2D codes require precise alignment and rely on extracting entire blocks, leading to inefficiencies and inaccuracies, especially when codes are rotated or tilted, limiting their application in augmented reality.
Innovation Solution
A method that removes peripheral blocks, sets weights for block selection based on recognition rates, and calculates rotation/tilt to quickly and accurately determine 2D codes by focusing on internal blocks with high visibility and partitioning them into groups for reduced computational processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional methods extract entire blocks in the captured 2D code image, then the recognition process can be performed, but the computational process becomes complicated and the recognition speed decreases
Solution Approach 1:
The patent segments the 2D code image into multiple blocks and further divides these blocks into different regions (first region, second region, third region) based on their positions. This segmentation allows the system to process only relevant blocks rather than entire blocks, reducing computational complexity and improving recognition speed.
Solution Approach 2:
The patent extracts only the necessary blocks from the captured 2D code image based on the determined tilt angle. By taking out only the relevant blocks that contain actual code information and excluding peripheral or irrelevant blocks, the system reduces the computational load while maintaining recognition accuracy.
2Adaptability or versatility
If the 2D code image is captured without precise alignment, then the code can be recognized in various orientations, but the conventional methods cannot determine the code information
Solution Approach 1:
The patent performs preliminary determination of the tilt angle of the 2D code image before extracting and processing the code blocks. By calculating the tilt angle in advance using the captured image data, the system can then correctly orient and extract the code blocks regardless of the image's initial orientation, enabling recognition of rotated or tilted codes.
3Productivity
If arbitrary blocks are assigned black or white colors when unclear, then the recognition process can proceed, but the accuracy of 2D code information extraction decreases
Solution Approach 1:
The patent applies different processing qualities to different blocks based on their positions and the determined tilt angle. Blocks in the first region (central area) are processed with higher precision to determine their colors, while blocks in other regions may be handled differently. This local quality approach ensures accurate color determination for critical blocks while maintaining overall process efficiency.
Data Source
AI summary
The present disclosure provides a method for recognizing 2D code information. A method for recognizing 2D code information according to the present disclosure, which is related to a method for recognizing 2D code information by capturing a 2D code image composed of rectangular cells, comprises obtaining a 2D code image using a mobile device equipped with a camera or an image capture device; performing a first data extraction process that identifies individual blocks storing a position pattern, an alignment pattern, and data from the obtained 2D code image; setting a first weight for removing outer blocks located outside a line connecting the position pattern and the alignment pattern with respect to a central block among the individual blocks of the 2D code image on which the first data extraction process has been performed; partitioning inner blocks located inside the position pattern and the alignment pattern among individual blocks of the 2D code image into a plurality of block groups composed of at least one or more inner blocks and setting a second weight for determining whether to select a block group according to a recognition rate of blocks constituting each block group; removing outer blocks set to the first weight and selecting a block group for which a weight has been set as having a high recognition rate among block groups set to the second weight; determining and recognizing a 2D code by comparing the selected block group with a block group of a pre-stored 2D code; and displaying an object to a user in the form of augmented reality based on the recognized 2D code information.