Detection of Fungal Contagion in Food Items Using Enhanced Image Segmentation
B. K. Mishra1, A. K. Rath2, P. K. Tripathy3
1Bikram Keshari Mishra*, Department of Computer Science and Engineering, Veer Surendra Sai University of Technology, Burla, India.
2Amiya Kumar Rath, Department of Computer Science and Engineering, Veer Surendra Sai University of Technology, Burla, India.
3Pradyumna Kumar Tripathy, Department of Computer Science and Engineering, Silicon Institute of Technology, Bhubaneswar, India.
Manuscript received on July 20, 2019. | Revised Manuscript received on August 10, 2019. | Manuscript published on August 30, 2019. | PP: 1748-1757 | Volume-8 Issue-6, August 2019. | Retrieval Number: F8434088619/2019©BEIESP | DOI: 10.35940/ijeat.F8434.088619
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Today the consumer demands for superior quality and safe food products. In order to obtain healthier products we need to emphasize on superior detection capabilities to identify any presence of foreign materials on them which are responsible for making them unhygienic. Image segmentation is one such technique which is vastly employed in such domains. It identifies the affected portion from the other regions. Hence, we made an effort to apply image segmentation to discover the existence of fungal contagion in food items. In this paper, an attempt has been made to use clustering as an approach in image segmentation. Few improved cluster-based image segmentation techniques like K-Means, MCKM, FEKM and FECA were used on quite a variety of food items to detect the existence of any kind of fungal growth on their surface. The results segmentation obtained were analyzed to verify their effectiveness by using few known performance measures including SC, RMSE, PSNR, MSE, MAE and NAE. The various food images were segmented to obtain both their gray scale and colored results. As per our anticipation, the outcome of FECA based segmentation is by far much sounder in contrast to the other methods. More or less every value of chosen quality measures offer encouraging results for FECA based segmentation technique as compared to the others, which implies accurate identification of fungal growth on food surfaces was achievable.
Keywords: FECA, FEKM, Fungus detection, Image segmentation, K-Means, MCKM.