Related Experiment Video
Updated: May 14, 2026
06:41
A New Technique for Quantitative Analysis of Hair Loss in Mice Using Grayscale Analysis
Published on: March 9, 2015
Hair detection in dermoscopic images using percolation.
Ana Afonso1, Margarida Silveira
1Institute for Systems and Robotics-Instituto Superior Tecnico, Av. Rovisco Pais, 1, 1049-001 Lisboa, Portugal. ana.afonso@ist.utl.pt
Summary
This study introduces an efficient percolation algorithm for detecting hair artifacts in dermoscopy images. The new method significantly outperforms existing DullRazor software in accuracy, improving diagnostic reliability.
Area of Science:
- Dermatology
- Medical Image Analysis
- Computer Vision
Background:
- Artifacts like hair obscure skin lesion details in dermoscopy images, hindering automated analysis.
- Accurate artifact detection and removal are crucial pre-processing steps for reliable diagnostic tools.
Purpose of the Study:
- To propose and evaluate an efficient algorithm for detecting hair in dermoscopy images.
- To compare the proposed hair detection method against the established DullRazor software.
Main Methods:
- Utilized an efficient percolation algorithm that analyzes pixel intensity and connectivity.
- Classified image points as hair based on the linear shape of connected pixel clusters.
- Validated the method on real-world dermoscopy images.
Main Results:
- The proposed percolation-based method demonstrated effective hair detection.
- Achieved over 10% improvement compared to DullRazor in reducing both false positives and false negatives.
- Indicated superior performance in identifying hair artifacts.
Conclusions:
- The novel percolation algorithm offers a more accurate and reliable approach to hair artifact removal in dermoscopy.
- This method enhances the quality of dermoscopy images for subsequent automated analysis.
- The findings suggest potential for improved diagnostic accuracy in skin lesion assessment.
