Title: Deep convolutional neural network for laser forward scattering image classification in microbial source tracking
Authors: Bin Chen
Addresses: Department of Electrical and Computer Engineering, Purdue University Northwest, Hammond, IN, 46323, USA
Abstract: The colony-based laser scatter imaging for microbial source tracking heavily relies on the power of optical scattering image classification. While carefully handcraft feature extraction achieved excellent results for the colonies with certain sizes for optimal classification results, the classification accuracy drops quickly for smaller or larger colonies outside of the colony size range. In this study, a deep convolutional neural network was implemented for laser scattering image feature extraction and classification. The results show that the deep learning classification method clearly outperforms the traditional clustering methods with high accuracy and consistency for host species with a wide range of colony sizes. It also provides comparable accuracy for the colonies with the optimal sizes.
Keywords: deep learning; convolutional neural network; microbial source tracking; laser imaging.
International Journal of Computational Biology and Drug Design, 2019 Vol.12 No.3, pp.261 - 267
Received: 23 Apr 2018
Accepted: 10 Jul 2018
Published online: 23 Jul 2019 *