Superpixel Segmentation Using Boundary Preserving Distance Metric
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional superpixel segmentation techniques are not well-suited for processing highly textured images, which are common, and lack computational efficiency and compactness in segmentation.
Innovation Solution
A computer-implemented method using a boundary preserving distance metric for superpixel segmentation, involving probabilistic boundary maps, connected component analysis, and adaptive region size determination, with the quick shift algorithm and Bhattacharyya distance, to generate accurate and compact superpixel segmentations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional superpixel segmentation techniques are used, then processing speed may be maintained, but segmentation accuracy for highly textured images deteriorates
Solution Approach 1:
The patent changes the parameter used for segmentation from conventional intensity-based metrics to a boundary-preserving distance metric that incorporates texture information. This allows the segmentation algorithm to adapt to highly textured images by considering boundary strength and texture characteristics in the distance calculation, thereby improving segmentation accuracy for textured regions.
Solution Approach 2:
The patent introduces a boundary map as an intermediary element that captures boundary information separately. This boundary map is then integrated into the distance metric calculation, allowing the segmentation process to preserve boundaries more effectively in textured images without directly modifying the core segmentation algorithm.
2Productivity
If conventional segmentation methods are used, then computational simplicity is maintained, but computational efficiency deteriorates
Solution Approach 1:
The patent performs preliminary computation of a boundary map before the main segmentation process. This boundary map pre-processes and stores boundary information that can be efficiently reused during segmentation, avoiding repeated boundary detection calculations and improving overall computational efficiency despite the added initial step.
Solution Approach 2:
The patent divides the segmentation process into distinct phases: boundary map computation, distance metric calculation using the boundary map, and final segmentation. This segmentation of the processing pipeline allows each phase to be optimized independently and improves computational efficiency by organizing operations in a logical sequence.
3Quantity of substance
If conventional segmentation is applied, then processing speed is maintained, but segmentation compactness deteriorates
Solution Approach 1:
The patent modifies the distance metric parameters to incorporate boundary preservation terms with weighted coefficients. By adjusting these parameters, the algorithm can produce more compact superpixels with fewer total superpixels while maintaining processing speed, as the boundary-aware metric guides merging decisions more effectively.
Data Source
AI summary
Systems and methods for implementing a superpixel image segmentation technique using a boundary preserving distance metric are disclosed. The disclosed technique segments a digital image into superpixels comprising contiguous pixel regions sharing similar characteristics. Superpixel image segmentation techniques presented herein utilize a boundary preserving distance metric. A boundary preserving distance metric presented herein measures the similarity between two pixels of a digital image at least partially based on a boundary probability values of the two pixels and surrounding pixels.