Scanning Electron Microscope Image Calibration with Improved Watershed Algorithm
A novel approach for calibrating scanning electron microscopy(SEM)images is introduced.We leverage the enhanced watershed algorithm to mitigate the challenges posed by intricate image features and shadow disruptions encountered across various detector modes.Initially,noise interference is mitigated while preserving genuine data integrity through preprocessing methodologies encompassing adaptive filtering and gradient transformations.Subsequently,in order to counteract shadow distortions within watershed markers,the adaptive enhancement technique is implemented to ensure marker acquisition universality across diverse detector configurations.Notably,chessboard sample assessments reveal superior performance of the proposed method when InLens and SE2 detectors are applied,and surpasses conventional calibration techniques across comprehensive evaluation metrics.Comparative analyses of calibration measurements against reference standards and manual assessments demonstrate a negligible relative error of within 7％,attesting to the method's exceptional accuracy and laying a robust foundation for the analysis of samples.
watershed transformationscanning electron microscopy(SEM)microanalysisimage processing
万方数据
维普
