--Your friends at LectureNotes

Note for Internet of Things - IOT by Dipu Saha

  • Internet of Things - IOT
  • Note
  • Electronics and Instrumentation Engineering
  • B.Tech
  • Uploaded 5 months ago
Dipu Saha
Dipu Saha
0 User(s)
Download PDFOrder Printed Copy

Share it with your friends

Leave your Comments

Text from page-3

International Journal Of Electrical, Electronics And Data Communication, ISSN: 2320-2084 http://iraj.in Volume-5, Issue-1, Jan.-2017 of services such as maps, videos via YouTube. We have developed a functional prototype to demonstrate our work. Overall, the prototype provides an easily extendable framework that can be utilized to provide even more functionality to the user. In our future work we will investigate how the surrounding context of the user and the environment can be utilized in order to provide optimal service experiences in the home environment.The system can be made much more useful to the users by adding more functionality like integrating light settings, speech processing, etc. V. FUNCTIONAL OVERVIEW The proposed mirror is designed to perform i.e.: Fig 2, several functionalities that can be summarized as follows: a) Mimic a natural mirror interface: b) A flat monitor is used for the mirror display. A one way mirror is used to provide real time display of what is located in front of the Smart Mirror using Raspberry Pi thereby mimicking the function of a regular mirror. c) Personalised Information services: Users will be able to obtain minute updates of latest news and public headlines, weather reports as well as get reports of our interests. d) Customized management of profiles: Users can create their own profiles and store them in the system. According to this profile, customized services are provided to the user. REFERENCES [1] Adobe Flex 2 http://www.adobecom/products/flex/; accessed: February 2007. [2] ERCIM Working Group SESAMI, Smart Environments and Systems for Ambient Intelligence. http://www.ics.forth.gr/sesami/. [3] Memory Mirror http://www.cc.gatech.edu/fcele/cl/projects/dejaVu/mnmjinde x~hml. [4] Philips Homelab. http:// www.research.philips. com/ technologies/misc/homelab/index.html [5] M. S. Raisinghani, A. Benoit, J. Ding. M. Gomez, K. Gupta, V. Gusila. D. Power, and 0. Schmedding. Ambient intelligence: Changing forms of human computer interaction and their social implications. Journal of Digital Information, 5(4), 2004. [6] F. Bomarius, M. Becker, and T. Kleinberger. Embedded intelligence for ambient-assisted living. ERCIM News, 67:19-20, 2006. [7] P.L. Emiliani and C. Stephanidis. Universal access to ambient intelligence environments: Opportunities and challenges for people with disabilities. IBM SystemsJournal, 44(3):605-619, 2005. [8] M. Friedewald, 0. Da Costa, Y. Punie, P. Alahuhta, andS. Heinonen. Perspectives of ambient intelligence in the home environment. Telematics and Informatics, 22(3):221-238, 2005. [9] Tatiana Lashina. Intelligent bathroom. In European Symposium on Ambient Intelligence (EUSAI'04), Eindhoven, Netherlands, 2004. [10] L. Ceccaroni and X. Verdaguer. Magical mirror: multimedia, interactive services in home automation. In Proceedings of the Workshop on Environments forPersonalized Information Access - Working Conferenceon Advanced Visual Interfaces (AVI 2004), pages 10-21, New York, NY, USA, 2004. ACM Press. Fig 2 Block Diagram CONCLUSION We have designed a futuristic smart mirror that provides natural interaction between users and the ambient home services. The mirror display is provided by a flat LED display monitor which displays all the necessary information which are useful for the user. The mirror also provides a picture-in-picture sub-display to facilitate the display  Design and Development of a Smart Mirror using Raspberry PI 65

Text from page-4

ISSN(Online): 2319-8753 ISSN (Print): 2347-6710 International Journal of Innovative Research in Science, Engineering and Technology (A High Impact Factor, Monthly, Peer Reviewed Journal) Visit: www.ijirset.com Vol. 7, Special Issue 2, March 2018 IOT Enabled Smart Mirror for Physically Challenged Vivek.V 1, Homan Rajan.G 2, Vijay.R 3, Surendar.S 4 , R. Geetha 5 U.G. Student, Computer Science and Engineering, S.A. Engineering College, Chennai, Tamil Nadu, India123 Professor, Computer Science and Engineering, S.A. Engineering College, Chennai, Tamil Nadu, India4 ABSTRACT: Technology has been improved to live people’s life most easily. Developing an application in such a way that the technologies reach the human to live a sophisticated life. Smart Mirror is one of the best ways to bridge a gap between the human and technologies. Since each person use the Mirror in his daily life, the mirror provides a natural means of interaction through which the residents can control the household smart appliances and access personalized services. So building time management technique in mirror will be useful and an efficient one. Making the smart mirror helps the physically challenged people much by making them up to date with latest technologies and provide security to them. Existing smart mirror are meant to be an entertaining people and for business purpose. So we proposed to build a smart mirror in order to help the physically challenged people to secure them and the busy people to effectively manage their time in their daily routines. I. INTRODUCTION Effective time management is one of the most import ant factors for every person’s life, especially if the person is physically challenged. By increasing the integration of technology in our lives helps us to maintain an efficient schedule. Keeping the event and the social media notification up to date is made easier through technology such as PC’s, smart phones etc. But using these gadgets doesn’t make us to use in busy life and also provide distractions that can interrupt anyone’s routine. So concept of integrating those high-tech features in mirror helps user to effectively manage their time in their daily routine. This high-tech feature helps the physically challenged people. They cannot move from one place to another often for their every need. So voice recognition will be very helpful by the way of controlling the Home appliances. Security option will make them to be safe in home where physically challenged need no other security appliances. In an emergency case, if they need to contact any one for help, just by voice commanding. Emergency alerting system using GSM benefits the users for up-to-date notification from mirrors. The various technologies used in smart mirror are Artificial Intelligence, Deep learning, Cloud. Artificial intelligence has extended its intelligent conclusion retrieval is being achieved using Machine learning. It derives the conclusion based upon the collected data. Machine learning is latest technology of making the machine to learn the data and to bring the conclusion by making itself a decision. Deep learning is the growing technique of determining the conclusion with high accuracy. Deep learning is a kind of machine learning in which ML learns without being explicitly programming. Once the program is implemented it should react based on real time data in variable environment. But Deep learning uses convolutional learning and clustering learning to get deeper results what machine learning does. Deep learning uses layering concepts in which it contains input layer , output layer and the intermediate hidden layers. Deep learning is a class of machine learning algorithms that multiple layers are cascaded of nonlinear processing units for feature extraction and transformation. Each successive layer uses the output from the previous layer as input; Copyright to IJIRSET www.ijirset.com 334

Text from page-5

ISSN(Online): 2319-8753 ISSN (Print): 2347-6710 International Journal of Innovative Research in Science, Engineering and Technology (A High Impact Factor, Monthly, Peer Reviewed Journal) Visit: www.ijirset.com Vol. 7, Special Issue 2, March 2018 Learning is done in both supervised and unsupervised manners, learning the multiple levels of representations that correspond to different levels of abstraction. Layers that have been used in deep learning include hidden layers of artificial neural networks. Types of Deep learning models are: Unsupervised Pertained Networks (UPNs), Convolutional Neural Networks (CNNs), Recurrent Neural Networks, Recursive Neural Networks. Convolution neural network, as a kind of depth learning model, can automatically learn and extract features from data. Its generalization ability is superior to traditional methods, and has been successfully applied to pattern classification, object detection and object recognition [1]. Here Conventional Neural network is used and Convolutional Neural Networks (CNN) has made great success on image and audio, which is the important component of deep learning [1]. Using the development of computer vision, pedestrian detection in the intelligent auxiliary driving, intelligent monitoring, pedestrian analysis and intelligent robot, and other fields have been widely used. Hinton and their students in the 2012 ILSVRC using convolution neural network got first place, and the classification task Top - 5 error rates is 15.3%, which is much better than the others [1].CNN will learn to recognize patterns across space. So, as you say, a CNN will learn to recognize components of an image (e.g., lines, curves, etc.) and then learn to combine these components to recognize larger structures (e.g., faces, objects, etc.). Recurrent Neural Network will similarly learn to recognize patterns across time. So a RNN that is trained to translate text might learn that "dog" should be translated differently if preceded by the word "hot". Cluster learning is a concept of collective learning network which predict and detect object based on collective data from various layers Test set accuracy comparison for a convolutional neural network (CNN) and a clustering learning (CL) network with 1 and 2 layers on the SVHN dataset [2]. And also the results in the CIFAR10 dataset in figure 2. As in the SVHN case, we compared results of accuracy in the test set for 4 cases: clustering learning with 1 layer (CL 1 layer), clustering learning with 2 layers (CL 2 layers), a 1layer and a 2-layers convolutional neural network (convnet 1l, 2l) [2]. Table 1 : Deep learning Comparison Layer 1 layer 2 layers Best Worst References Convolutional Neural Network 57% 50% [1] Clustering learning 52% 38% [2] Convolutional Neural Network 70% 45% [1] Clustering learning 40% 36% [2] From the table 1, we can conclude that convolution neural network is an efficient one. In Existing system, the smart mirror is developed based on three objectives: [3] 1. To design a prototype Smart Mirror using Raspberry PI 2. To develop a voice recognition system to facilitate the implementation of Smart Mirror. 3. To carry out the testing process on Raspberry PI for usability evaluation to users. But the security and home safety is not being implemented, [3] The author focused on developing smart mirror for smart life where bring an Human Technology interface and to meet the functional requirements. There are predefined available commands to enhance to user to use those commands to use smart mirror efficiently.[3] Copyright to IJIRSET www.ijirset.com 335

Text from page-6

ISSN(Online): 2319-8753 ISSN (Print): 2347-6710 International Journal of Innovative Research in Science, Engineering and Technology (A High Impact Factor, Monthly, Peer Reviewed Journal) Visit: www.ijirset.com Vol. 7, Special Issue 2, March 2018 Touch less data acquisition is achieved [4] using Vision camera and Multispectral camera for emotional analysis. The Wize Mirror features an advanced sensing framework for unobtrusive acquisition of videos, images, and 3D scans of individuals standing in front of the mirror. Multispectral cameras are used for analysis the skin tissues and microcirculation. In [4] the main objective is to promote a healthy lifestyle using smart mirror. In the paper [5], the author proposed a Wize mirror that uses multisensory cardio-metabolic risk monitoring system. In this paper, mirror is built as a result of the FP7 funded SEMEOTICONS (Semeiotic Oriented Technology for Individuals Cardio metabolic risk self-assessment and Self-monitoring). In this, user can self-monitor their well-being status over time and to improve their lifestyle using latest technology. In the paper [6] , the author proposed a smart mirror to monitor the elder at the home, it collect the data like pulse , emotion, Blood pressure of them and collect those data in EC system, This system then process the data with the threshold limit set in the EC system , When a threshold limit exceed , and alert messages with the critical value of forwarded to the elder care taker via SMS In the paper [7] , the author developed a smart mirror that is an reflective interface, which displays the news feed, calendar, reminder features to the user . Author also proposed it with an face recognition system that identifies the user, Here no other features are being proposed. In the paper [8], the author proposed an interactive mirror with Home aware, It consist of many interactive features like Load cell, RFID, webcam. The load cell is used to weigh the user and get his BMI, RFID is used to detect the garments worn by the user, and webcam is to perform the face recognition. Here no home automation and security features are being proposed. In the paper [9], the author developed a smart mirror which consist of the Automated home control system , It uses the Relay module to control the appliances. No other security features and social media interaction is being proposed In the paper [10], the author designed and developed a mirror that has an user identification module, Interactive games for the user and Activities tracking. It enhanced with the emotion detection. Here it is not implemented for any security features. Table 2: Comparison for existing project with our proposal S.No 1. Title Smart mirror for smart life 2. A smart mirror to promote a healthy lifestyle 3. Wize Mirror - a smart, multisensory cardiometabolic risk monitoring system Copyright to IJIRSET Output Effective Voice user interface and interact using available commands, Home control system is implemented. Reduce the socio-economic burden of chronic and widespread diseases, such as cardiovascular and metabolic diseases. Translate the semeiotic code of the human face into computational descriptors and measures, automatically extracted from videos, multispectral images Cons Security feature and Home safety is not being implemented. No social media activities Security feature and Home safety is not being implemented, No voice user interface. References [3] No social media activities. No Home control system. [5] www.ijirset.com [4] 336

Lecture Notes