SCT-Center Color Segmentation for Melanoma Diagnosis

A computer vision application for medical image processing that segments skin lesions using the SCT-Center (Spherical Coordinate Transform) algorithm to assist in melanoma diagnosis.

📋 Overview

This project implements a 5-step color segmentation algorithm specifically designed for analyzing melanoma images. The SCT-Center transformation decouples color information from brightness variations, making it robust for medical image analysis under different lighting conditions.

🎯 Features

• Median Filtering: Noise reduction with configurable kernel size
• SCT-Center Transformation: Advanced color space transformation (15×15 quantization = 225 colors)
• Interactive ROI Selection: Point-and-click interface for region of interest selection
• Morphological Operations: Automatic noise removal and shape refinement
• Area Calculation: Precise measurement of segmented regions with percentage calculations
• Real-time Visualization: Live preview of segmentation results

🚀 Algorithm Steps

Step 1: Preprocessing

• Applies 3×3 median filter to reduce acquisition noise
• Optional color quantization from 2²⁴ to 256 colors

Step 2: SCT-Center Color Segmentation

• Converts RGB to spherical coordinates (L, angle A, angle B)
• Maps to 15×15 quantized color space (225 total colors)
• Decouples color information from brightness variations

Step 3: Interactive Fine-tuning

• User clicks on regions of interest
• Creates RGB cubic clusters around selected points
• Adjustable delta parameter for cluster size (0-128)
• Real-time mask updates with blue overlay

Step 4: Morphological Filtering

• Erosion followed by dilation (opening operation)
• Uses 3×3 cross-shaped structural element
• Removes small artifacts while preserving main structure
• Generates binary mask and color overlay

Step 5: Area Computation

• Calculates segmented area in pixels
• Computes percentage relative to total image area
• Displays results for medical analysis

🛠️ Technical Requirements

Dependencies

• OpenCV 4.x - Computer vision library
• Visual Studio - C++ development environment
• Windows OS - Required for file dialog functionality

Build Instructions

1. Clone the repository:

git clone https://github.com/yourusername/sct-center-melanoma-segmentation.git
cd sct-center-melanoma-segmentation

2. Open project in Visual Studio
3. Configure OpenCV paths in project settings
4. Build and run the application

📊 Usage

1. Launch Application: Run the executable
2. Select Image: Choose a melanoma image file through the dialog
3. View Preprocessing: Observe median filtering results
4. Interactive Segmentation:
   • Click on regions of interest in the SCT-Center image
   • Adjust Delta slider for sensitivity (recommended: 15-25)
   • Continue clicking until entire lesion is covered
5. Review Results:
   • Binary mask shows segmented region
   • Overlay shows original image with highlighted ROI
   • Area and percentage statistics are displayed
6. Process Next Image: Press any key to continue with another image

🎨 Visual Results

Step-by-Step Results

Step 1: Preprocessing

Before	After
Original melanoma image	After 3×3 median filtering

Step 2: SCT-Center Transformation

15×15 quantization producing 225 distinct colors

Step 3: Interactive Segmentation

Real-time ROI selection with adjustable delta parameter

Steps 4-5: Morphological Filtering & Results

Binary Mask	Color Overlay
ROI in black	Blue overlay on original

📈 Medical Applications

This tool assists dermatologists and medical professionals in:

• Lesion Boundary Detection: Precise segmentation of irregular melanoma shapes
• Area Measurement: Quantitative analysis for medical documentation
• Color Analysis: Identification of variegated coloring patterns
• Diagnostic Support: Objective measurement tools for clinical assessment

🔬 Algorithm Advantages

• Lighting Independence: SCT transformation handles varying illumination
• Color Preservation: Maintains critical color information for diagnosis
• User-Guided: Interactive selection ensures medical expertise integration
• Morphological Refinement: Automatic cleanup of segmentation artifacts
• Quantitative Output: Precise measurements for medical records

📝 Research Background

Based on the SCT-Center algorithm developed for variegated coloring identification in skin tumors. The spherical coordinate transformation provides superior color-brightness decoupling compared to traditional RGB analysis, making it particularly suitable for medical imaging applications.

⚠️ Important Notes

• This tool is designed for research and educational purposes
• Not intended for clinical diagnosis without professional medical supervision
• Results should be validated by qualified dermatologists
• Accuracy depends on image quality and proper user interaction

🤝 Contributing

Contributions are welcome! Please feel free to submit issues, feature requests, or pull requests.

📚 References

[1] Umbaugh Scott E., "Computer Vision and Image Processing", Prentice Hall, NJ, 1998.

📧 Contact

For questions or collaborations, please open an issue on GitHub.

---

Keywords: Melanoma, Medical Imaging, Color Segmentation, SCT-Center, Computer Vision, OpenCV, Skin Cancer, Image Processing