Smart prevention system for combating pandemic situations
Dr.j.s. leena jasmine1, annie getzial.j2, irene bennita.b3, muthu ramya.r4, priyadharshini.s5 Department of electronics and communication engineering
Velammal engineering college, chennai
Article History: Received: 11 January 2021; Revised: 12 February 2021; Accepted: 27 March 2021; Published
online: 4 June 2021
Abstract
During this coronavirus pandemic, hand hygiene has become the foremost important measure to curb the outspread of the disease as instructed by world health organization using soaps, alcohol-based hand sanitizers etc. Even if anyone cleans their hands, they have to touch the surface which would be contaminated by virus. Corona virus would last on a contaminated surface from several hours to days depending upon environmental conditions. In addition to hand hygiene face masks have also been used by the people in public places to protect them from the virus. Wearing facemasks is very important but nowadays either people are not wearing masks or wearing masks improperly. In this paper a technique has been developed which will detect the people’s faces and separates them into people who are wearing masks and people who are not wearing masks or improper wearing using deep learning and image processing using matlab .also a smart hand sanitizer is designed with contactless temperature reader with entrance door control for covid safety using atmega328p microcontroller, solenoid lock ,infrared sensor, temperature sensor which will help to provide a solution to the hurdles faced by the security guards in emphasizing the people in hand sanitizing action. Some people may enter the building without sanitizing their hands. So, the automatic hand sanitizer dispenser is kept at the entry door and the door is controlled by the dispenser. Before entering the premises, a temperature sensor will detect the body temperature of the person. With this dispenser, an infrared sensor is attached to sense the presence of hands. After the sanitizer is dropped in the hands, the door will unlock and the door opens. If the condition is not satisfied, the door remains locked.
Keywords: covid-19, deep learning, image processing, matlab Introduction
Government of india is taking all necessary steps to corroborate that the people are prepared well to accept the hurdles and warnings created by the growing pandemic of covid-19 the corona virus. With active uphold of the people of india, the growth of virus has been suppressed in our country. The foremost important think about preventing the spread of the virus locally is to authorize the citizens with the proper information and taking precautions as per the recommendations furnished by ministry of health & family welfare. A variety of the precautionary measures include cleaning or washing hands at frequent intervals, wearing a mask, social distancing,
Monitoring health regularly. These measures are given as recommendations by the govt which should be followed at public places.
There are more ways with which the covid-19 can spread from person to person which include airborne transmission, surface transmission etc., during this pandemic situation it's
Counseled by who to take care of healthy sanitizing habits, temperature screening and far more.
Since the breakout of covid-19, facemasks have been used by the people in public places to protect them. Besides the patients who are suspected of covid-19 infection are advised to wear masks to curb the spread of virus, the healthy persons are also advised to wear mask to protect them. The masks when worn properly can protect us but the efficacy of wearing masks has now been decreased because of improper wearing of masks. The advancements in computer vision and deep learning has resulted in lot of opportunities to develop various things. One major model of dnn is the convolution neural network which is now used in computer vision field on a large scale. On proper training with various dataset images, convolution neural network can be able to classify the images into face mask wearing and not wearing images.
In [1], the authors developed a three-category classification method for facemask wearing condition like facemask wearing, no facemask wearing and improper facemask wearing. In [2] the authors presented various deep learning algorithms for identification and detection. In [3], the ingman proposed a deep learning technique such as convolution neural network used in image processing for facial recognition. In [4], the author proposed a model which provides an integration of machine learning and deep learning methods. In [5], the authors have proposed an algorithm for facial image tracking and recognition using matlab and arduino for security and video surveillance purposes. In [6], the authors have put forward a system for automatic opening of doors using facial recognition. In [7], abhinandan sarkar has proposed a system of automatic hand sanitizer dispenser system with a contact temperature detector. In [8], the authors have proposed a system which has an automatic sanitizer dispenser with a door control connected to it. In [9], the author has presented an automatic hand sanitizer dispenser with a non-contact temperature reader. In [10], the authors proposed a touchless completely programmed sanitizer with an inbuilt ultrasonic sensor which is useful for detecting the hands and the sanitizer gets poured onto the hands. In [11], the authors have presented an automatic tap for hand-wash which is handy and is easy to add in water pipe for hand – wash routine. In [12], the author has proposed a contactless infrared thermometer which can measure the temperature and time of acquisition. It has a measurement error less than 0.5c and a liquid crystal display which displays the temperature and time. In [13], the authors have proposed a smart medication dispenser which consist of medication scheduler and dispenser controller. In [14], the authors have proposed an smart hand sanitizer system which is useful with variety of containers which is concentrated on utilizing the elasticity of pumps and enhancing people’s approach to devices. In [15], the authors proposed a technology for face recognition by interfacing matlab with arduino which uses the theory of image acquisition and fundamentals of digital image processing. In [21] the authors have proposed a face mask detection system with convolution neural network and yolo framework.
In the existing system, a security guard was stationed at the entrance to take temperature checks, face mask wearing checks and to sanitize the customers entering the shop or any building. Thermal screening was done using a temperature gun. Then a foot operated sanitizer machine was installed to allow people to sanitize on their own. The bottle would dispense the liquid by pressing the pedal with foot. A separate contactless temperature screening was set up which uses infrared technology.
Therefore, in this paper a technique has been executed which will check the face mask wearing condition automatically using image processing and deep learning techniques such as cnn. This process involves four steps: getting input image, image pre-processing, convolution neural network, and face mask wearing condition identification. Also, a smart hand sanitizer dispenser is designed which controls the door. This dispenser uses an infrared sensor to detect the presence of hands. Together with this, a temperature sensor will check the body heat of the person. After the utilization of sanitizer and temperature sensing the door will unlock and also the lcd display is provided to display the message. This automated technology will help to cut back the spread of this novel virus.
Methodology
This system comprises four modules. They are facial mask detection, smart hand sanitizer dispenser, contactless temperature detector and a door controller.
Figure 1. Proposed system
First all the persons who are entering the building have to pass through a face mask detection. For this purpose, image processing (dataset images) and deep learning are used to check whether the person is wearing a mask or not.
deep learning is a set of techniques for learning with the help of neural network. Neural network is inspired programming paradigm which allows computer to learn from data there are many learning algorithms. One of these is the cnn (convolution neural network). This cnn is used to process data in multiple layers of arrays. This is mainly used for image or facial recognition applications. The major difference between cnn and other neural networks is that this cnn takes and operates on the 2d images directly than concentrating on feature extraction like other neural networks.
The algorithm used in this system is cnn (convolution neural network). It is a kind of feed forward neural network made up of several layers. If a bunch of pictures of faces are fed to the system, they acquire knowledge about some fundamental things like edges, dots, dark spots, bright spots and this gets identified in the first layer.the second layer include those that are identifiable like eyes, nose, mouth and the third layer recognize those that look like faces. The convolution neural network includes an image which is a two-dimensional array of pixels and by using that cnn checks whether one is wearing mask or not by following various steps like filtering, convolution, pooling, normalization, rectified linear units (relus) and finally fully connected layer. By using this technique, the input image is compared with the stored dataset images in matlab and the output is obtained. This facemask wearing condition output is fed into the arduino uno.
Figure 2.cnn
The cnn algorithm receives an input image that goes through the layers to identify features and acknowledge the image and then it brings about the classification result. The main steps involved in the programming part are getting the image, preprocess them and to get the classifier trained by inputs which is the modified image. Many functions are utilized in the programming part such as randomly changing the light of the image and padding is used to retrieve the size of the image. A dataset is used here which contains face images with masks and images without masks. This dataset is used for convolutional neural network training. Before this preprocessing is done. Here image is resized into 100 x 100. Now the trained dataset images are included into the convolutional layer. Figure 3 shows the resized image.
Figure 3 input image and the resized image(100x100) INPUT IMAGE
PREPROCESSING
CNN
Convolutional layer is the major building block of cnn, which is useful for feature detection. Every convolutional layer applies a new filter to the convoluted image of the previous layers that can extract another feature. If more filters are used, the more feature the cnn can take out from the image. Figure 4 depicts all the dataset images which has been taken for training.
Figure 4. Dataset
After every convolutional layer, which is a linear function, a relu activation function is applied. It establishes non-linearity in the convolution layer.
Pooling layer is used to eliminate some unnecessary information, which can decrease the calculation by less parameters. Max pooling layer is used. Fully connected layer reduces the affect to features which is caused by pixel position when removing from image.
After this a graph has been plotted to predict the accuracy and loss. Over a series of epoch, the model is able to distinguish the images.
Figure 5 depicts the images passing through the layers like convolution layer, pooling layer
The facemask wearing and not wearing condition is checked through the above techniques and the result is given as an input to the arduino uno.
Next the persons have to sanitize their hands. For this an infrared sensor is used for the detection of any motion or a human being. When the infrared sensor detects the presence of hand, the arduino receives the reading and activates the servo motor. If it detects the hand within 10cm, the motor moves from 0° to 180°, the sanitizer is poured out on the hand from the spray pump and it will wait for 2 seconds to return to 0°. Then the people have to pass through a temperature check. For this the arduino is attached with a temperature sensor which detects the body temperature of the person and displays the temperature in lcd display in celsius. If the temperature is above the normal temperature, then it will display ‘body temperature is high’ in the lcd display and the door will not open. If the face mask is worn, hands are sanitized and the temperature is normal, the solenoid lock will deenergize, ‘the door is open’ message is displayed on the lcd display and the door will be opened. Then the door will automatically close after 10 seconds. Figure 6 depicts the flow chart of the entire system
Figure 6 flow chart
Experimental results
The facemask wearing condition is checked using matlab (dataset images) using image processing and deep learning techniques such as cnn. Figure 7 shows the facemask wearing condition and figure 8 shows the no
facemask wearing condition.
Figure 7 facemask detection condition
Figure 8 no facemask detection condition
If the temperature is above the normal temperature, then the lcd will display ‘body temperature is high’ and the door will open and “the door is open’ message is displayed on the lcd display.
Figure 9 shows the model of the proposed system. Figure 10 shows the temperature detection which shows the body temperature of the person. Figure 11 depicts the door control operation.
Figure 10. Temperature detection
figure 11. Door control
Conclusion
This paper is very helpful in maintaining hand hygiene without any hurdles to enter into any premises. This is much safer because of its contactless detectors which reduces the possibility for contamination. This is a handy system which can be used by anyone easily. This paper can be very helpful in the battle with an invisible enemy and can act as a weapon for surviving in this pandemic situation.
References
1. Bosheng qin and dongxiao li, “identifying facemask -wearing condition using image super-resolution with classification network to prevent covid-19, pp 20-5236,2020
2. s. A. Hussain, a.s.a.a. balushi, “a real time face emotion classification and recognition using deep learning model”, j. Phys.: conf. Ser. 1432 (2020) 012087, doi: 10.1088/1742- 6596/1432/1/01208 3. xin, m., wang, y. “research on image classification model based on deep convolution neural network”.
J image video proc. 2019, 40 (2019)
4. Amrit kumar bhadani, anurag sinha, “a facemask detection using machine learning and image processing techniques”, an international journal on engineering science and technology, november 2020
5. Maneesh ayi, ajay kamal ganti, maheswariadimulam, badigantikarthik, manishabanam, g.vimala kumari, “face tracking and recognition using matlab and using arduino”, vol 5, issue 12, pp 2321-9653, 2017. 6. Hiteikhtarlwin, aungsoekhaing, hlamyo tun, “automatic door access system using face recognition”,
international journal of scientific and ,2015
7. X. Jia, "image recognition method based on deep learning," 2017 29th chinese control and decision conference (ccdc), chongqing, 2017, pp. 4730-4735, doi: 10.1109/ccdc.2017.7979332
8. Image classification using convolutional neural networks deepikajaswal, sowmya.v, k.p.soman
9. I.yugashini, s. Vidhyasri, k. Gayathri devi, design and implementation of automated door accsessing system with face recognition, international journal of science and modern engineering (ijisme), volume-1, issue-12, november 2013.
10. Ayushi gupta, ekta sharma, nehasachan and neha tiwari. “door lock system through face recognition using matlab”. International journal of scientific research in computer science and engineering, vol1, issue-3, 30 june 2013.
11. Abhinanandhan sarkar, “design of automatic hand sanitizer with temperature sensing”, international journal of innovative science and research technology, vol.5, issue 5, pp 2456-2165,2020
12. Enerstedozie, watimbajanat and zainakalyankolo, “design and implementation of a smart hand sanitizer dispenser with door controller using atmega328p”, international journal of engineering and information system 4, issue 6, pp 2643-640x,2020
13. Marlon gan rojo, jolanbacaysy, eunelfa regie calibara and wubishetdegife, “non-contact temperature reader with sanitizer dispensor (nctrsd)”, international journal of scientific and reasearchpublications, vol 10, issue 9, pp 2250-3152,2020.
14. Rakshith and dr.k.b.shiva kumar, “a novel automatic sanitizer dispensor”,international journal of engineering reasearch and technology,pp 2278-0181,2020.
15. P.h.tsai,c.y.yu,c.s.shih and j.w.s.liu, “smart mediactiondispensor” , ieee systems journal, vol 5, issue 1,2011
16. Eddyyusuf,m.n.mohammad,ismailidaood,sitihumarahkamarulbahrain,s.al-zubaidi,omar ismael al-sanjary,sairah a.k. “2019 novel coronavirus disease(covid-19):smart contactless hand sanitizer-dispensing system using iot based robotics technology”,vol 29,pp 215-220,2020
17. Hurriyatul fitriyah*, edita rosana widasari, eko setiawan, and brian angga kusuma, “interaction design of automatic faucet for standard hand-wash” matec web of conferences (2018)
18. G. Long, “design of a non-contact infrared thermometer,” int. J. Smart sens. Intell. Syst., vol. 9, no. 2, pp. 1110–1129, 2016, doi: 10.21307/ijssis2017-910
19. Juhui lee,jin-young lee and kwang gi kim, “design of automatic hand sanitizer system compatible with various containers”, healthc inform res,vol 23, issue 3,pp 243-247,2020.
20. Kiran r, m.lohith,yogesh e,sai kumar a,anitha j, “interfacing of matlab with arduino for face recognition”, international journal of science,technology and engineering 6th national conference by tjit,2017
21. Yahia said, pynq-yolo-net: an embedded quantized convolutional neural network for face mask detection in covid-19 pandemic era, (ijacsa) international journal of advanced computer science and applications, vol. 11, no. 9, 2020.
22. Yadav, shashi. "deep learning based safe social distancing and face mask detection in public areas for covid-19 safety guidelines adherence." international journal for research in applied science & engineering technology (ijraset). 2020.
23. loey, mohamed, gunasekaran manogaran, mohamed hamed n. Taha,and nour eldeen m. Khalifa. "a hybrid deep transfer learning model
24. With machine learning methods for face mask detection in the era of the covid-19 pandemic." measurement (2020): 108288
25. [24] t subhamastan rao , s anjali devi , p dileep , 26. m sitha ram” a novel approach to detect face 27. mask to control covid using deep
28. learning” european journal of molecular & clinical medicine issn 2515-8260 volume 07, issue 06, 2020 29. [25] md. Shahriar islam, eimdadul haque moon, md. Ashikujjaman shaikat, mohammad jahangir alam, “a novel approach to detect face mask using cnn”, third international conference on intelligent sustainable systems, pp 978-1-7281,2020
30. [26] l. Wen, x. Li and l. Gao, "a transfer convolutional neural network for fault diagnosis based on resnet-50", neural comput. Appl., vol. 32, no. 10, pp. 6111-6124, 2020.
31. [27] t. Ojala, m. Pietikainen and t. Maenpaa, "multiresolution gray-scale and rotation invariant texture classification with local binary patterns", ieee transactions on pattern analysis and machine intelligence, vol. 24, no. 7, pp. 971-987, july 2002.
32. [28] m.k.j. khan, n. Ud din, s. Bae and j. Yi, "interactive removal of microphone object in facial images", electronics, vol. 8, no. 10, 2019.
33. [29] mahesh t. Dubey, vaibhav p. Kale, prajwal p.jagtap, ankita p. Mende, vrushabh p. Kalbande, kirti b. Nagne, “ iot based automatic hand sanitizer dispenser” , international research journal of engineering and technology, vol 7,issue 12,pp 2395-0056, dec 2020
34. [30] krizhevsky, i. Sutskever, and g. E. Hinton, “imagenet classificationwith deep convolutional neural networks,” in advances in neural informationprocessing systems 25, f. Pereira, c. J. C. Burges, l. Bottou, andk. Q. Weinberger, eds. Curran associates, inc., 2012, pp. 1097–1105.
35. [31] k. He, x. Zhang, s. Ren, and j. Sun, “deep residual learning for imagerecognition,” 2016 ieee conference on computer vision and patternrecognition (cvpr), pp. 770– 778, 2016
36. [32] jay p. Sipani, riki h patel, trushit upadhyaya, “ temperature and humidity monitoring and control system based on arduino and sim900a gsm shield” , international journal of electrical,electronics and data communication,vol 5, issue 11, pp 2320-2084, nov 21017
37. [33] deeksha srivastava, awanish kesarwani, shivani dubey, “ measurement of temperature and humidity by using arduino tool and dht 11”, international research journal of engineering and technology, vol 5, issue 12, pp 2395-0072, dec 2018
38. [34] s. Kalaiarasi , shubham gautam, anshuman behera 39. ,mansi mewara, “ arduino based temperature
40. and humiditysensor”,journal of network
41. communication and emerging technologies ,vol 8, issue4 42. , pp 2395-5317, april 2018
43. [35] kamweru paul kuria , owino ochieng robinson 44. mutinda mutava gabriel, “monitoring temperature 45. and humidity using arduino nano and module- 46. dht11 sensor with real time ds3231 data logger 47. and lcd display”,ijert, vol 9, issue 12,dec 2020
