A Hybrid Fuzzy based Cross Neighbor Filtering (HF-CNF) for Image Enhancement of fine and coarse powder Scanned Electron Microscopy (SEM) images

• Published in: Journal of intelligent & fuzzy systems 2022, Vol.42 (6), p.6159-6169
• Main Authors: Jayaseelan, Samuel Manoharan, Gopal, Sakthivel Thirumalai, Muthu, Sangeetha, Selvaraju, Sivamani, Patel, Md Saad
• Format: Article
• Language: English
• Subjects: Blurring; Coarseness; Fault detection; Fineness; Flaw detection; Image enhancement; Image filters; Image processing; Random noise; Scanning electron microscopy; Scientific papers
• ISSN: 1064-1246; 1875-8967
• DOI: 10.3233/JIFS-212561

Image enhancement is one of the most critical stages towards any image processing application. The outcome of image enhancement determines the accuracy and precise nature of the overall output from the image processing under interest. This research paper has shown specific interests towards enhancement of Scanned Electron Microscopic (SEM) images which are primarily concerned with projection of fine details exist in internal details of surfaces, metals, fine powders, fibers etc. These fine details play a dominant role in detection of minute cracks, artifacts, progressing faults, texture of powders, their coarseness or fineness, internal details of fibers in forensics. However, due to the image capturing process which is through conventional camera-based models, noise tends to be a major source in degrading or blurring the underlying vital information. A cross neighbor fuzzy filter is a hybrid combination called Hybrid Fuzzy Based Cross Neighbor Filtering (HF-CNF) which is proposed in this research paper in order to minimize impulse and random noise to a great extent also to fine tune the further processing stages. The proposed method has been subjected to extensive analysis by comparison with state of art and recent benchmark methods and superior performance justified in terms of several validation metrics.