Curvature-Based Active Region Segmentation for Improved Image Processing of Aspergillus Species

Authors

  • Nur Rodiatul Raudah Mohamed Radzuan Department of Mechatronic Engineering, Faculty of Electrical Engineering and Technology, Universiti Malaysia Perlis, 02600 Arau, Perlis, Malaysia
  • Haryati Jaafar Department of Mechatronic Engineering, Faculty of Electrical Engineering and Technology, Universiti Malaysia Perlis, 02600 Arau, Perlis, Malaysia
  • Farah Nabilah Zabani Department of Mechatronic Engineering, Faculty of Electrical Engineering and Technology, Universiti Malaysia Perlis, 02600 Arau, Perlis, Malaysia
  • Fatin Norazima Mohamad Ariff Department of Mechatronic Engineering, Faculty of Electrical Engineering and Technology, Universiti Malaysia Perlis, 02600 Arau, Perlis, Malaysia
  • Fatin Nadia Azman Fauzi Department of Mechatronic Engineering, Faculty of Electrical Engineering and Technology, Universiti Malaysia Perlis, 02600 Arau, Perlis, Malaysia

DOI:

https://doi.org/10.37934/araset.46.1.157174

Keywords:

Image segmentation, Active region, Level set function

Abstract

Aspergillus is one of the most ubiquitous of the airborne saprophytic fungi that can withstand various climatic conditions and could cause multiple type of illness. It can be beneficial to humankind and also can be infectious to humans and animals. Direct microscopic is used by trained microscopist as one of the alternatives in identification process to any specimen that suspected of having fungal infection. Confirmation towards identification is often necessary as the structure of Aspergillus is complex and dissimilar in each cycle. In addition, the structure of some species of Aspergillus are the almost same, which can be incorrectly recognized. In prevention of misidentification, computer-based Aspergillus species identification is proposed. The detection process is the earliest and important process hence, this paper proposed an active region-based segmentation method in order to detect the presence of fungi. This method is literally not depending on the gradient or sharp edges of the object and implementing level set function for curve evolution which able to reduce the computational cost. Originally, this function was developed for tracking fluid interfaces but in this study, this function has been applied to fungi database. Two different methods were tested and compared to observe their ability to segment different 80 of Aspergillus images which included four species. Experiments conducted have been compared with the baseline technique and the proposed method is outperformed in terms of accuracy, specificity with average of 90% and PSNR value of greater than 40dB. Meanwhile the active contour (snake) was slightly underperformed but well performed particularly in terms of sensitivity with greater than 80% for all the species. Moreover, upon scrutinizing the dice coefficients provided in both tables, it becomes apparent that there is a lack of significant variance in the values, except in the instance of Aspergillus fumigatus (active region-based) that which produces a result below 36%.

Downloads

Download data is not yet available.

Author Biographies

Nur Rodiatul Raudah Mohamed Radzuan, Department of Mechatronic Engineering, Faculty of Electrical Engineering and Technology, Universiti Malaysia Perlis, 02600 Arau, Perlis, Malaysia

raudahradzuann@gmail.com

Haryati Jaafar, Department of Mechatronic Engineering, Faculty of Electrical Engineering and Technology, Universiti Malaysia Perlis, 02600 Arau, Perlis, Malaysia

haryati@unimap.edu.my

Farah Nabilah Zabani, Department of Mechatronic Engineering, Faculty of Electrical Engineering and Technology, Universiti Malaysia Perlis, 02600 Arau, Perlis, Malaysia

farahzabani@gmail.com

Fatin Norazima Mohamad Ariff, Department of Mechatronic Engineering, Faculty of Electrical Engineering and Technology, Universiti Malaysia Perlis, 02600 Arau, Perlis, Malaysia

fatin.azima.ariff@gmail.com

Fatin Nadia Azman Fauzi, Department of Mechatronic Engineering, Faculty of Electrical Engineering and Technology, Universiti Malaysia Perlis, 02600 Arau, Perlis, Malaysia

fatinnadiaaf@unimap.edu.my

Downloads

Published

2024-04-29

How to Cite

Nur Rodiatul Raudah Mohamed Radzuan, Haryati Jaafar, Farah Nabilah Zabani, Fatin Norazima Mohamad Ariff, & Fatin Nadia Azman Fauzi. (2024). Curvature-Based Active Region Segmentation for Improved Image Processing of Aspergillus Species. Journal of Advanced Research in Applied Sciences and Engineering Technology, 46(1), 157–174. https://doi.org/10.37934/araset.46.1.157174

Issue

Section

Articles

Most read articles by the same author(s)