Klasifikasi Citra Penggunaan Masker Menggunakan Multilayer Perceptron Backpropagation
Total View This Week0
Total View This Week0
Institusion
Institut Teknologi Telkom Purwokerto
Author
Azharuddin, Subhi
Subject
T Technology (General)
Datestamp
2022-04-14 07:14:47
Abstract :
During the current COVID-19 pandemic, regulations are enacted where everyone is required to wear a mask and perform physical distancing when outside the home. This is a new habit that will be familiarized to the Indonesian people by the government. So that this disciplined habit of using masks in public places can run well, a Classification System for Using Mask Image Using Multi-Layer Perceptron (MLP) was created. So that the environment such as companies or schools can discipline the regulations that have been made by the government by wearing masks when leaving the house. Retrieval of data in the form of image data to be taken with the dataset used. Determining the number of nodes in the hidden layer correctly affects the performance of the MLP on the mask image classification system. The image data will be processed using the computational library numpy, scikit-learn. And using the Multilayer Perceptron Classification (MLP) as the method, so that the level of accuracy obtained will be maximized. From the classification experiment using the MLP method, with a dataset of 1000 images used. The image is divided into 2 parts, namely 500 images of people wearing masks and 500 images of people not wearing masks. The design of the mask image classification system has been successfully completed in this study. The best testing results were obtained using the MLP classification and the sigmoid activation function had an accuracy of 70%. Keywords: Classification, Image, Mask, MLP, Accuracy
During the current COVID-19 pandemic, regulations are enacted where everyone is required to wear a mask and perform physical distancing when outside the home. This is a new habit that will be familiarized to the Indonesian people by the government. So that this disciplined habit of using masks in public places can run well, a Classification System for Using Mask Image Using Multi-Layer Perceptron (MLP) was created. So that the environment such as companies or schools can discipline the regulations that have been made by the government by wearing masks when leaving the house. Retrieval of data in the form of image data to be taken with the dataset used. Determining the number of nodes in the hidden layer correctly affects the performance of the MLP on the mask image classification system. The image data will be processed using the computational library numpy, scikit-learn. And using the Multilayer Perceptron Classification (MLP) as the method, so that the level of accuracy obtained will be maximized. From the classification experiment using the MLP method, with a dataset of 1000 images used. The image is divided into 2 parts, namely 500 images of people wearing masks and 500 images of people not wearing masks. The design of the mask image classification system has been successfully completed in this study. The best testing results were obtained using the MLP classification and the sigmoid activation function had an accuracy of 70%. Keywords: Classification, Image, Mask, MLP, Accuracy
Download
book
BibTex
Latex, Jabref
cloud_download
Original Resource
url resource Institution
