AN IMPROVED METHOD FOR AUTOMATIC CHARACTERIZATION OF ALUMINUM OXIDE NANOPORE FESEM IMAGES

The application of nanopores in nanotechnology majorly depends on their customized synthesis and appropriate characterization. Since ages, the characterization of these pores has been a major challenge. The researchers, chemist and nanotechnologist have used manual methods and software like ImageJ in the characterization of the nanopores, which require human intervention. In such methods, measuring every nanopore in the sample image is time consuming and cumbersome task. In the proposed study an attempt has been made to develop an automated tool using digital image processing techniques to overcome the disadvantages associated with the good old manual characterization. This automated method has employed five different segmentation techniques to segment the nanopores from the alumina before their characterization and compared the geometrical features; wall thickness and nanopore size results obtained from all these segmentation techniques with the manual results provided by the chemist. The proposed study depicts that the nanopores acquired by global thresholding and active contour segmentation technique have a least average computational error of 0.28% as compared to k-means, region growing and watershed method which show 0.72%, 8.28% and 2.43% average computational error respectively and proves that global thresholding and active contour are the most appropriate segmentation techniques to extract the pores from nanoporous FESEM images. The average circularity of the pores in the images is ranging from 0.54nm to 0.86nm.