Photonics Systems Informatics
Statistical processing and machine learning based analysis for measurement obtained from photonics instrument
Data Processing
- Noise characterization and processing of Laser Induced Breakdown Spectroscopy (LIBS) measurement of scrap metal, skin tissue samples etc.
- Processing of LIBS measurement to compensate for undesired degradation in the signal due to experimental mechanisms
Pattern Classification
- Dimensionality reduction of LIBS spectra by selecting most informative spectral lines for accurate classification
- Machine learning based classification of scrap metal, skin tissue sample etc. using weighted averaged spectral images and soft decision of clusters synthesized by weighted sum of regression models
Systematic hardware based signal processing for high speed classification of sample based on prior information
Pre-clustering based Regression for Soft Classification
- Clustering based on correlation of element concentration from prior information
- Soft decision based clustering synthesized by the weighted sum of regression models
Hardware Accelerated Optical Signal Processing
- Field programmable gate array (FPGA) accelerated compressive sensing
- Machine intelligence investigation for optical measurements
