Digital Forensics and Internet of Things. Группа авторов

Figure 1.2 Layers of the model.

1.3.1 Neural Network

      A neural network is an instrument that is designed to model in the similar way in which the brain responds or executes a task or function; it is usually simulated in digital computer-based software or carried out by using electronic components. It can resemble the brain in the following aspects:

       • The knowledge is obtained by the network from its surrounding with the help of a learning procedure.

       • Interneuron link strength, known as synaptic weight, is used to accumulate the obtained knowledge.

       • The process that is operated to execute the learning procedure is known as the learning algorithm; the purpose of which is to reform the synaptic weights of the network in a well-organized mode to accomplish the desired layout objective.

       • It is also possible to improve its own topology.

       • Neural network is also mentioned in literature as neurocomputers, connectionist network, and parallel distributed processor.

       • Neural network attains its computing power at the beginning from its power of computer at first from the massively side-by-side distributed arrangement and next from its potential to learn and then generalize.

       • Generalization leads to the neural network constructing logical outputs for inputs not encountered throughout training (learning).

      An ANN is specified by the following:

       • Neuron model: Data processing component of the neural network.

       • An architecture: A group of neurons along with connections connecting neurons.

       • A training algorithm: It is used for instructing the Neural network by changing the weights to model a selected training task correctly on the instructing examples.

      1.3.2 Application of Neural Network in Face Recognition

      Face recognition implies comparing a face with the saved database of faces to recognize one in the given image. The associated process of face detecting is directly relevant to recognizing the face as the images of the face captured must be at first analyzed and then identified, before they get recognized. Face detection through an illustration assists to focus on the database of the system, improving the systems speed and performance.

      Artificial Neural Network is used in face recognition because these models can imitate the neurons of the human brain work. This is one of the foremost reasons for its role in face recognition.

1.4 Methodology

      1.4.1 Face Recognition

Face acknowledgment is subject to the numerical features of a face and is probably the most natural approach to manage face affirmation. It is one of the first robotized face affirmation structures. Marker centers (position of eyes, ears, and nose) were used to build a component vector (distance between the centers, point between them). The affirmation was performed by ascertaining the Euclidean distance between included vectors of a test and reference picture. Such a technique is vigorous against changes in enlightenment by its temperament; however, it has an immense disadvantage: The precise enlistment of the marker focuses is confounded, even with cutting edge calculations. Probably, the most recent work on mathematical face recognition was reported by Mulla M.R. [20]. A 22-dimensional component vector was utilized and it was investigated that huge datasets have appeared, that mathematical highlights alone may not convey sufficient data for face Recognition. Figure 1.3 depicts the detailed structure of face recognition system.

Figure 1.3 Structure of face recognition system.

      1.4.2 Open CV

      OpenCV (Open-Source Computer Vision) is a famous library developed by Intel in 1999. This platform has various libraries. It helps in real-time image processing and includes various algorithms. It is equipped with programming interface to various languages like C++, C, and Python.

      OpenCV 2.4 has a very useful new face recognizer class for face recognition.

      The currently available algorithms are as follows:

       • Eigenfaces (createEigenFaceRecognizer())

       • Local Binary Patterns Histogram (createLBPHFaceRecog-nizer())

       • Fisher faces (createFisherFaceRecognizer() )

      1.4.3 Block Diagram

This framework is controlled by Raspberry Pi circuit. Raspberry Pi electronic board is worked on battery power supply and remote web availability by utilizing USB modem; it incorporates camera, PIR movement sensor, LCD, and an entryway, as shown in Figure 1.4. At first approved countenances get enlisted in the camera. Then, at that point, the confirmation happens. At whatever point the individual comes before the entryway, PIR sensor will detect the movement; LCD screen shows the necessary brief and the camera begins perceiving the face; it perceives the face, and on the off chance that it is enrolled, it opens the entryway; if the face is not enlisted, then it will raise a caution and snaps an image and sends it on the qualifications. This is the means by which the framework works.

Figure 1.4 Block diagram of face recognition system.

      1.4.4 Essentials Needed

      SD card with 16GB capacity preinstalled with NOOBS.

      For display and connectivity:

      Any HDMI/DVI monitor or TV can be used for pi Display. HDMI cables will also be needed.

      Keyboard and mouse: wireless will also work if already paired.

      Power supply: USB cables can be used for this. Approximately, 2 A at 5 V will be needed to power the Raspberry Pi.

      Make an account on iotgecko.com for authentication check.

      1.4.5 Website

      If a person is unidentified, then a picture of is captured and sent to the website. All the monitoring data is sent over the website iotgecko.com so that the user can see the system status from anywhere and help boost the security.

      1.4.6 Hardware

Figure 1.5 depicts the components used:

       • Raspberry Pi 3 Model B+

       • Camera

       • Multimedia Mobile AUX System

       • PCB

       • 16X2 LCD Display

       • DC Motor