Changelog¶
All notable changes to the Hyperseed project.
The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.
0.1.0-alpha.3 - 2025-01-29¶
Added¶
- Initial release of Hyperseed - Hyperspectral Seed Image Analysis Tool
- Core functionality for hyperspectral image processing and analysis
- Support for ENVI format hyperspectral data files
- Comprehensive preprocessing pipeline with multiple methods:
- Standard Normal Variate (SNV)
- Smoothing (Savitzky-Golay, Gaussian, Moving Average, Median)
- Baseline correction (Polynomial, Rubberband, ASLS)
- Derivative computation (1st, 2nd order)
- Normalization (Min-Max, Max, Area, Vector, Peak)
- Multiplicative Scatter Correction (MSC)
- Detrending
- Advanced segmentation algorithms for seed detection:
- Threshold-based segmentation (Otsu, adaptive, manual)
- Watershed segmentation
- Connected components analysis
- Combined approach for robust detection
- Spectral extraction and analysis capabilities:
- Extract mean spectra from segmented regions
- Statistical analysis of spectral signatures
- Export to CSV and HDF5 formats
- Reflectance calibration with white and dark references
- Command-line interface (CLI) with multiple commands:
analyze: Full pipeline analysissegment: Segmentation onlybatch: Batch processingconfig: Configuration managementinfo: System information- Comprehensive test suite with ~80% code coverage
- Visualization tools for segmentation results
- Support for batch processing of multiple datasets
- Configuration system with presets (minimal, standard, advanced)
- Validation metrics for segmentation quality (IoU, Dice, F1-score)
Features¶
- Performance: Optimized for large hyperspectral datasets
- Flexibility: Modular architecture allowing custom pipelines
- Extensibility: Plugin-ready architecture for custom algorithms
- Documentation: Comprehensive documentation and examples
- Testing: Extensive test coverage ensuring reliability
Dependencies¶
- Python >=3.10
- NumPy, SciPy, scikit-learn, scikit-image
- OpenCV for image processing
- Pandas for data manipulation
- Matplotlib for visualization
- Click for CLI interface
- Rich for enhanced terminal output
Known Issues¶
- Some empty module directories reserved for future features
- Visualization requires display environment (not suitable for headless servers without proper configuration)
Contributors¶
- Nishad Thalhath - Lead Developer
- Deepa Kasaragod - Contributor
Note¶
This is an alpha release (alpha.3). The API and features are still under active development and may change in future releases.