Region Recognition Using Boundary Curves and Contour Maps
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current intelligent surveillance technologies and image processing methods can only determine a rough region where a target object is located, leading to reduced accuracy in surveillance.
Innovation Solution
A region recognition method that involves obtaining a scene image, determining a specified boundary curve, generating a contour map based on this curve, and identifying a target region within the scene, allowing for more precise region detection by handling curved boundaries and features.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional image processing methods are used to determine the region where a target object is located, then the surveillance system can identify a rough region, but the range of the region is not accurate, reducing the accuracy of surveillance
Solution Approach 1:
The patent segments the region recognition process into multiple distinct steps: obtaining scene image, determining boundary curve, generating contour map, and determining target region. This segmentation allows each step to be optimized independently, improving overall accuracy while managing complexity through modular processing
Solution Approach 2:
The patent introduces boundary curves with curvature to define region boundaries instead of simple straight lines. The boundary curve determination step uses curved geometric representations to accurately model complex scene boundaries, significantly improving region recognition accuracy for irregularly shaped regions
2Measurement precision
If simple straight line boundaries are used for region determination, then the method is simple to implement, but the accuracy of region recognition is reduced
Solution Approach 1:
The patent replaces simple straight line boundaries with boundary curves that can accurately represent complex geometric shapes. The boundary curve is defined by multiple control points and parameters, enabling precise representation of curved and irregular boundaries while maintaining a systematic implementation approach
Solution Approach 2:
The patent performs preliminary boundary curve determination and contour map generation before final target region identification. This preliminary processing establishes an accurate geometric framework that simplifies the subsequent target region determination step, separating the complexity of boundary definition from the simplicity of region identification
Data Source
AI summary
A region recognition method includes obtaining a scene image of a scene to-be-recognized; determining a specified boundary curve of the scene image; generating a contour map corresponding to the scene image according to the specified boundary curve; and determining a target region in the scene according to the contour map.