Browsing by Author "Dr. K. M. A. Salam"

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Embargo
    Covid-19 detection from X-Ray images using deep learning methods with a web app
    (North South University, 2022) Saadman Sakib; Nafisa Hassan; Sifat Anwar; Dr. K. M. A. Salam; 1632440042; 1912043642; 1812887642
    The World Health Organization (WHO) classified COVID-19 a global epidemic in 2019. COVID-19 is caused by SARS-CoV-2, also known as the severe acute respiratory syndrome coronavirus-2, which was found in China in late December 2019. The entire planet had been affected within a few months. COVID-19 has infected millions of individuals all over the world, making it a global health issue. The disease is usually contagious, and those who are infected can readily transfer it to others. As a result, monitoring is an effective way to stop the virus from spreading. Another condition caused by a virus linked to COVID-19 is pneumonia. The severity of pneumonia can range from minor to life-threatening. This is especially harmful for children, the elderly, and individuals with health problems or impaired immune systems. We employed deep transfer learning to classify COVID-19 and pneumonia in this experiment. Because there has been so much research on this subject, the suggested strategy focuses on enhancing accuracy and employs both a transfer learning methodology and a custom-made model. Different pretrained deep convolutional neural network (CNN) models were used to extract deep features. Classification accuracy was used to evaluate performance to a considerable extent. According to the findings of this study, deep transfer learning can detect COVID-19 and pneumonia from CXR images. VGG-16 was 94 percent accurate in pretrained custom models, InceptionV3 was 96 percent accurate, ResNet50 threshold was 83 percent accurate, and Xception was 92.82 percent accurate. All of these models are correct, however InceptionV3 is the most accurate one for detection of Covid-19. A lightweight python framework, Flask is used to incorporate our algorithm and build our entire web application and it may be widely applicable in health sector. The information is then extracted and the results of Covid-19 detection are displayed.
  • Thumbnail Image
    Item
    Embargo
    The future of industrial protection and monitoring automation system in the Era of the Internet of Things
    (North South University, 2019-04-30) Sabbir Hassan Suvo; Washim Sarker Shuvo; Fahim Hossain Talukdar; Md. Nahid Enam; Dr. K. M. A. Salam; 1512421643; 1511540643; 1420925043; 1320536043
    In the 4th industrial revolution, automation of industry is increasing day by day. For connecting with automation Internet of things (IoT) is a new technological revolutionary name. IoT is playing a massive role in this era to monitor industry properly and also for quick protection automatically. The key idea of our project is to protect the industry from quick damage and also warning the user by ioT server. Besides, it will also prevent the resources and people’s life from unwanted damages. This was also our another priority to make people’s life more secure who works in those industries. In this project, we used thingspeak built in iot server. When sensor can detect any problem, our Arduino Uno microcontroller will upload the data to the server through wi-fi and server will show the warning update of that sensor automatically. With this our another output will go into 20*4 Lcd display which will show the update of 4 sensord continuously. The system was designed in such a way so that it can take care need of industry and give industries protection system a new life