Natural Feature Point Marker Detection in Image Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image processing technologies face challenges in seamlessly integrating markers into image data without causing discomfort, as they often require visible identifiers like QR codes, which can disrupt the natural appearance of scenes.
Innovation Solution
An image processing device that accepts image data containing circular features as markers, using the positional relationship and diameter ratio of registered perfect circles to identify and specify markers, allowing for natural feature points to be used without the need for explicit identifiers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If markers with identifiers or feature information are attached to enable image processing, then marker recognition accuracy is improved, but the natural appearance of the image is degraded
Solution Approach 1:
The invention extracts the essential geometric features (circular shape, position, diameter) from traditional markers and uses only these extracted features for recognition. This removes the need for attached identifiers or feature information while maintaining recognition capability, thus preserving the natural appearance of the image.
Solution Approach 2:
The invention uses perfect circle templates as reference models to compare against circles in the image. By copying the ideal geometric properties of perfect circles (exact circularity, known diameter relationships) and using them as comparison standards, the system achieves accurate marker detection without requiring additional marker elements that would disrupt the natural image appearance.
2Reliability
If traditional markers with identifiers are used, then marker specification reliability is improved, but user comfort and image integration are degraded
Solution Approach 1:
The invention enables circles in the image to serve themselves as markers through their inherent geometric properties. Natural circular features in the image automatically provide the necessary recognition information (position, diameter, circularity) without requiring additional marker elements or identifiers, thus eliminating user discomfort while maintaining reliable specification.
Solution Approach 2:
The invention changes the recognition parameters from traditional marker properties (identifiers, patterns, colors) to pure geometric parameters (circularity, diameter, position). This parameter transformation allows reliable marker specification using only the natural geometric properties of circles in the image, eliminating the need for disruptive marker attachments.
3Shape
If markers are integrated naturally without identifiers, then image naturalness is improved, but marker detection difficulty increases
Solution Approach 1:
The invention performs preliminary preparation by establishing perfect circle templates with known geometric properties before the actual detection process. These pre-prepared reference circles serve as comparison standards, making the detection of natural circles in the image more straightforward by providing a clear reference for matching geometric parameters.
Solution Approach 2:
The invention replaces the mechanical/visual system of reading identifiers or recognizing patterns with an automated geometric parameter comparison system. By substituting manual or complex pattern recognition with automated comparison of mathematical parameters (circularity, diameter ratios, positions), the system maintains image naturalness while reducing detection difficulty through computational efficiency.
Data Source
AI summary
An image processing device according to the present application includes an accepting unit and a specifying unit. The accepting unit accepts image data that includes therein at least two diagrams that can be extracted as circles. The specifying unit specifies, by checking a parameter that is calculated from a pair of two circles included in the image data accepted by the accepting unit against a parameter that is calculated from a pair of two perfect circles that have previously been registered as markers that are used for image processing, the pair of the two circles included in the image data as the markers.