Abstract—When matrices on images of the sea bottom is classified, it takes human cost with human visual classification. It is necessary to develop a method of an automatic classification of the matrices. In the past, we proposed a method by using Convolutional Neural Network (CNN). The classification is performed using information surrounding the attention pixel selected randomly for the classification. However, the influence of irrelevant information on surrounding the attention pixel is considered to be one of the causes of the misclassification. In this paper, we performed a method of blurring away from the attention point, and showed usefulness of the method through experiments.
Index Terms—Quadrat image, convolutional neural network, patch image, Gaussian filters.
Akiyuki Sakai is with University of the Ryukyus, Graduate School of Science and Engineering, Okinawa 903-0213 Japan (e-mail: k198542@eve.u-ryukyu.ac.jp). Shinya Nozaki is with University of the Ryukyus, Faculty of Engineering, Okinawa 903-0213 Japan (e-mail: nozaki@tec.u-ryukyu.ac.jp). Takashi Sakamaki is with Tohoku University, Graduate School of Engineering, Miyagi 980-8579 Japan (e-mail: takashi.sakamaki.a5@tohoku.ac.jp).
[PDF]
Cite:Akiyuki Sakai, Shinya Nozaki, and Takashi Sakamaki, "Classification of Matrices in Quadrat Images by Using Convolutional Neural Network and Gaussian Filter," Journal of Advances in Computer Networks vol. 8, no. 1, pp. 10-13, 2020.
Copyright © 2020 by the authors. This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited (
CC BY 4.0).
