Seed Point Determination via Voting and Density Maps
Solution Overview
Problem
Manual image segmentation is time-consuming and challenging, especially when images contain numerous features with varying shapes, and existing automatic methods struggle with accurate seed point identification.
Innovation Solution
A method and apparatus for determining seed points in images by creating a voting map, density map, and consolidating local maxima based on pixel voting and weighting, which helps in automatically segmenting images by identifying reliable seed points for feature extraction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual segmentation is performed, then segmentation accuracy can be maintained, but time consumption increases significantly
Solution Approach 1:
The system enables automatic segmentation by having the image data itself provide voting information from multiple pixels to determine seed points, eliminating the need for manual intervention while maintaining accuracy through collective pixel voting and density map analysis
Solution Approach 2:
The patent introduces intermediate processing structures (voting map, density map, and candidate seed point identification) that mediate between raw image data and final segmentation results, enabling automated accurate segmentation without direct manual input
2Loss of time
If automatic segmentation methods are used, then time consumption is reduced, but seed point identification accuracy deteriorates
Solution Approach 1:
The patent combines voting information from multiple pixels into a unified voting map, then merges this with density map analysis to identify candidate seed points, achieving accurate automatic seed point identification through combined information processing
Solution Approach 2:
The system uses feedback mechanisms where the voting map and density map provide iterative refinement of seed point candidates, with local maxima identification and path analysis feedback ensuring accurate seed point selection automatically
3Adaptability or versatility
If features have widely varying shapes, then image complexity increases, but seed point determination becomes more problematic
Solution Approach 1:
The patent applies local quality analysis by examining pixel-level voting patterns and density distributions in different regions of the image, allowing accurate seed point determination for features with varying shapes through localized analysis rather than global assumptions
Solution Approach 2:
The patent transitions from analyzing only image intensity to incorporating spatial voting dimensions and density map layers, adding dimensional information that enables accurate seed point identification for features with diverse shapes by considering multiple spatial relationships
Data Source
Figure 1~4
Figure 2
Figure 3
AI summary
Embodiments of the present invention provide a computer-implemented method of processing image data for determining seed points with respect to the image data, comprising determining a voting map, wherein said voting map is based on voting locations determined for at least some pixels within the image data, determining a density map, wherein determining said density map comprises determining a weighting for at least some pixels within the voting map, wherein the weighting is determined as a non-linear function based on a value of each of the at least some pixels of the voting map, and determining the seed points in the image data based on the density map.