| A rapid and cost effective technique for identification and classification of microorganisms was developed using fluorescence microscopy, image analysis and neural networks. Microorganisms were stained using two fluorescent dyes---diamidino-2-phenyl-indole (DAPI) and Acridine Orange. After staining the microorganisms with fluorescent dyes, images of the microorganisms were captured using a CCD camera attached to an Olympus BX60 light microscope. Geometrical, optical and textural features were extracted from the images of microorganisms using image analysis. Optical parameters were extracted from the gray level histogram of the images of microorganisms. Textural features were extracted using histogram based techniques, co-occurrence matrix, run length algorithm and discrete wavelet transform. From the geometrical, optical and texture parameters extracted from images of microorganisms, the best identification parameters that could classify the microorganisms with higher accuracy were selected using a probabilistic neural network (PNN). PNN was then used to classify the microorganisms with a 100% accuracy using those identification parameters. |
