QR Code Spatial Tracking via Positioning Block Feature Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current AR technologies using two-dimensional codes as markers face issues of slow recognition speed and low tracking accuracy due to unstable two-dimensional code contours and the time-consuming process of regenerating feature point sets for new codes.
Innovation Solution
A method and system that utilize main positioning blocks of two-dimensional codes, specifically QR codes, for feature detection, where central points, corner points, and boundary edge points are used as a universal feature point set for accurate and efficient recognition of spatial location information, eliminating the need for repeated regeneration of feature point sets.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If two-dimensional code contour method is used, then tracking accuracy can be high for similar contours, but tracking accuracy becomes low for significantly different contours due to pattern variations
Solution Approach 1:
The patent extracts the main positioning blocks from the two-dimensional code structure to form a dedicated feature point set. By isolating and utilizing only the positioning block features (corners and edges) rather than the entire code contour, the system achieves consistent tracking accuracy across different code patterns while maintaining reliability.
Solution Approach 2:
The patent creates a universal feature point set based on positioning blocks that can be applied to all two-dimensional codes regardless of their specific patterns or values. This universal approach allows the same feature extraction method to work reliably across diverse code types, solving the problem of varying tracking accuracy for different contours.
2Measurement precision
If regeneration method is used to generate preset feature point sets, then tracking accuracy can be maintained, but processing speed becomes slow due to repeated regeneration steps
Solution Approach 1:
The patent performs preliminary extraction of feature points from positioning blocks during the code design or initialization phase, creating a preset feature point set in advance. This preliminary action eliminates the need for repeated regeneration during actual tracking operations, significantly improving processing speed while maintaining accuracy.
Solution Approach 2:
The patent uses the positioning block structure as a template or copy that can be universally applied across different two-dimensional codes. Instead of regenerating feature points from scratch for each code, the system copies the established positioning block feature set, which maintains accuracy while dramatically reducing processing time.
3Adaptability or versatility
If standard two-dimensional code contours are used as feature points, then the system can work with existing codes, but the feature points become unstable due to different code values and patterns
Solution Approach 1:
The patent applies local quality by focusing exclusively on the positioning blocks, which have consistent structural characteristics regardless of the overall code pattern. By extracting features only from these specific local regions (corners and edges of positioning blocks), the system achieves stable feature points while maintaining compatibility with various code types.
Data Source
Figure 1
Figure 2A
Figure 2B
AI summary
A method and a system for recognizing location information in a two-dimensional code are provided. The method comprises: acquiring a two-dimensional code in an image (S102); performing, according to a main positioning block of the two-dimensional code, feature detection to recognize location information in the two-dimensional code (S104); and determining, according to the location information in the two-dimensional code, spatial location information of the two-dimensional code in the image (S 106). Through the method, an augmented reality solution in which a two-dimensional code serves as a marker has favorable tracking performance.