Related Experiment Video
Updated: Dec 23, 2025
12:27
Large-scale Reconstructions and Independent, Unbiased Clustering Based on Morphological Metrics to Classify Neurons in Selective Populations
Published on: February 15, 2017
7.2K
A Local Neighborhood Robust Fuzzy Clustering Image Segmentation Algorithm Based on an Adaptive Feature Selection
1Changchun Institute of Optics, Fine Mechanics and Physics, Chinese Academy of Sciences, Changchun 130033, China.
Sensors (Basel, Switzerland)
|April 26, 2020
Summary
This study introduces an improved fuzzy clustering algorithm that enhances image segmentation performance, particularly in noisy conditions. The new method effectively suppresses noise and improves operational efficiency for remote sensing applications.
Area of Science:
- Computer Vision
- Image Processing
- Machine Learning
Background:
- Existing fuzzy local information C-means (FLICM) algorithms struggle to account for varying feature impacts on clustering segmentation.
- This limitation hinders accurate image segmentation, especially in the presence of diverse noise types.
Purpose of the Study:
- To propose a novel local fuzzy clustering segmentation algorithm that incorporates feature selection.
- To enhance the robustness and efficiency of image segmentation algorithms under various noise interferences.
Main Methods:
- Developed a local fuzzy clustering algorithm integrating a feature selection Gaussian mixture model.
- Incorporated spatial distance constraints into the local information function and introduced feature saliency into the objective function.
- Utilized Lagrange multiplier method for optimization and added neighborhood weighting to membership degree iterations.
Main Results:
- The improved FLICM algorithm demonstrated superior performance in segmenting images corrupted by Gaussian, salt-and-pepper, multiplicative, and mixed noise compared to original FLICM and FCM_S.
- Significant improvements were observed in peak signal-to-noise ratio and error rate metrics.
- The enhanced algorithm showed reduced iteration time and fewer iterations for convergence.
Conclusions:
- The proposed algorithm significantly enhances noise suppression capabilities, especially under strong interference.
- The improved efficiency facilitates remote sensing image acquisition in challenging, noisy environments.
- This work contributes to the development of robust anti-noise fuzzy clustering algorithms.
