International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 09 Issue: 11 | Nov 2022
p-ISSN: 2395-0072
www.irjet.net
FACE MASK DETECTION AND COUNTER IN THINGSPEAK WITH EMAIL ALERT SYSTEM FOR COVID19 ERIKI ANIRUDH SAI1,MIKKILINENI PHANISHWAR2 1ERIKI
ANIRUDH SAI &VIT VELLORE UNIVERSITY, ANIRUDHSAI868686@gmail.com PHANISHWAR & VIT UNIVERSITY,MIKKILINENIPHANISHWAR@gmail.com ---------------------------------------------------------------------***--------------------------------------------------------------------2MIKKILINENI
Abstract - Due to this unexpected pandemic, we are going
image and extracts the ROI of each face. This ROI is applied to each face mask classifier to detect the mask. Finally, it gives the output
these days, wearing a face mask became mandatory to save ourselves as well as others from the virus. But it is difficult to monitor every customer whether he is wearing a mask or not. But it is very important. So, to overcome this problem we came up with a solution to monitor every customer using a deep learning concept. So, we are developing a face mask detector with OpenCV/Keras and updating the count of customers waiting outside with an email alert system if at least one customer is not wearing a mask. This helps us easily identify the customers wearing masks or not, which helps us to take safety measures according to it. We tried using different types of platforms such as mobilev2net and resnet architecture but the accuracy of resnet architecture is more compared to the other architecture ( Size 10 & Italic , cambria font)
Key Words:
1.1 OBJECTIVE We aim to design a hands-free entry system using a face mask detector in surveillance to combat the further spread of the virus. This will ensure to reduce the transmission of pathogens on high-touch surfaces, like door handles, and to prevent entry in a community area without a face mask. To accomplish this task, we will fine-tune the Resnet architecture which will help us train hundreds of layers quickly and make sure that there won’t be a drop in the training percentage.
1.2 PROPOSED SYSTEM
Mask detection, OpenCV, E-mail,
Thingspeak
We will take an input from Realtime video as an input and this video is processed using the algorithm developed by us. We use the testing algorithm to detect if the person identified in the image is wearing a mask or not this then gives a digital output which includes the status of the person and the accuracy of the output. This output will be displayed on the screen
1. INTRODUCTION As we all know that there is an ongoing pandemic of coronavirus disease 2019 (COVID- 19) which is accelerating day by day, self-protection is the only wayout which can be done by wearing masks. Given this current situation, our team decided to make a face mask detector with people counter and an alert system. The basic task at hand is to check whether the person is wearing a mask or not through an available image or video. So in this era of automation and artificial intelligence, we decided to come up with a project that isgoing to automate the process of face mask detection using open CV and deep learning thereby making the life of frontline warriors easy. Developing this face mask detector is not only the way but we should develop it in a more portable way that can be used in any area. So, to make it portable we are using a Resnet classifier. This is the only way to make this project portable and user- friendly. There are two phases in this face mask detector.Phase1: Train the face mask detector. Phase2: Apply face mask detector. Phase 1 is the basic step in which wewill train the project using the datasets and train it to vary between different types of images. This will be the testing and comparison set to the real images. In phase 2,the given image is compared with the stored dataset and gives us the original output of whether the person is wearing a mask or not. The detailed explanation is in phase 2 the face mask classifier is loaded from the disk, then the camera detects the
© 2022, IRJET
|
Impact Factor value: 7.529
1.3 LITERATURE REVIEW [1] The vision towards computers is changing every 13 months. During this change using computer technologyin the neuro-scientific field also came to light in which this face detection is also one of the parts. [2] Algorithms based on principal component analysis are the basis of numerous studies in neuroscience and algorithmic face detection literature. eigenvector is one of the main concepts in differentiating and grouping thedata to train the device. [3] Every face is treated as a twodimensional face rather than three-dimensional oneto make identifying easier. But these images can be identified in any colour but the image should not be tinted with green colour. [4] A view-based multiple-observer eigenspace technique is proposed for use in face recognition under variable poses. In addition, a modular eigenspace description technique is used which incorporates salient features such as the eyes, nose, and mouth, in an eigen feature layer. This
|
ISO 9001:2008 Certified Journal
|
Page 84