Related Experiment Video
Updated: Apr 30, 2026
06:49
Automated Analysis of Dynamic Ca2+ Signals in Image Sequences
Published on: June 16, 2014
16.5K
Automated adjustment of region-based active contour parameters using local image geometry
IEEE Transactions on Cybernetics
|April 29, 2014
Summary
This study presents an automated method for active contour (AC) parameterization, improving image segmentation quality and objectivity. The novel framework guides AC models using local geometry, reducing manual trial-and-error for better results.
Area of Science:
- Computer Vision
- Image Processing
- Computational Geometry
Background:
- Parameterization of active contour (AC) models is crucial for segmentation quality, objectivity, and robustness.
- Manual adjustment of AC parameters is time-consuming and lacks objectivity.
Purpose of the Study:
- To introduce a novel framework for automated adjustment of region-based AC regularization and data fidelity parameters.
- To improve the efficiency and reliability of image segmentation using AC models.
Main Methods:
- Developed a framework linking AC energy term weights to structure tensor eigenvalues.
- Encoded local geometry by mining orientation coherence in edge regions.
- Guided AC models based on edge orientation coherence.
Main Results:
- Successfully applied the automated adjustment framework to four state-of-the-art AC models.
- Demonstrated comparable segmentation quality to empirical parameter adjustment.
- Validated on diverse datasets including natural, textured, and biomedical images, and image restoration models.
Conclusions:
- The proposed automated parameterization framework enhances AC model performance.
- Eliminates the need for cumbersome and time-consuming manual parameter tuning.
- Offers a robust and objective approach to image segmentation.
