Automatic detection and segmentation of different solar features taken by PICARD/SODISM Satellite (dark regions (Sunspots), bright or white regions, active region (networks and faculae) using:
- Machine learning
- Region growing technique.
- Thresholding
Modified region growing technique to detect the regions of interest (RoI) FAR error rate and FRR error rate for active regions The detection performance is enhanced further using a combination of modified region growing and neural network (NN) technique which is trained on statistical features extracted from the RoI and non-RoI. Using this combination the FAR has dropped to 2% for active regions, and 4% for filaments.
!!! Images and mask should have same names
Scikit-image module for image segmentation https://towardsdatascience.com/image-segmentation-using-pythons-scikit-image-module-533a61ecc980
About FAR & FRR https://www.bayometric.com/false-acceptance-rate-far-false-recognition-rate-frr/
AI for segmentation https://missinglink.ai/guides/computer-vision/image-segmentation-deep-learning-methods-applications/