Arduino based smart home security design using biometric recognition

As a standard feature, smart homes have a lock door that requires a password to open. However, burglars can simply defeat this security measure, and it is impossible to identify the user or person who opens the lock door. The authors are interested in building a system with the name research design of an Arduino-based smart home security system employing biometric recognition based on the aforementioned issues. The purpose of this study is to identify the performance requirements and design of an Arduino-based biometric home security system. This study is an engineering study that tries to identify the performance and design requirements for an Arduino-based smart home security system. The mechanical shape of the tool was created using the performance specification. The properties of the sensor, as well as the accuracy and precision of the tool, serve as the basis for the design specifications. The precision value of each sensor, namely the fingerprint sensor, is extremely good, while the precision values of the PIR sensor and magnetic sensor are 99.56% and 97.87%, respectively. The accuracy value of the sensor on the tool, primarily the fingerprint sensor, is very good.


Introduction
Technology plays a part in the construction of dream homes in the modern period, where technical advancements are occurring at a very rapid pace, in order to improve the comfort and safety of the home's residents [1].Homes today come with systems made of sensors that can monitor and regulate things like lights, temperature, alarms, and other household facilities.A smart home [2] is a home with this system installed.Researchers have been working on security system technology up until this point in order to provide technology that can ensure or boost a sense of security in the living environment.The most crucial component of a smart home is its security system [3]- [4].Security systems including alarm systems, fire alarms, and gas leak monitors are typically included in smart home.The lock door that requires a password to open is typically included in smart home.However, thieves can simply defeat this security measure, and it is impossible to identify the user or person who opens the lock door.
In [5] fingerprint-based door security that makes use of a fingerprint sensor.It has not been possible to provide faces or photographs of users of the technology in this study.Human biometrics must be used in the design of the security system in order to solve this issue.Through the use of specialized technology called biometrics, people can be recognized by their physical traits, which can be seen or detected.While using this biometric method, traits can be recognized in the form of fingerprints, facial shape, eyes, and voice [6].Human fingerprints will be the biometric used.The fingerprint sensor, ESP32 Cam video module, Arduino Uno, buzzer, door lock solenoid, magnetic sensor, and PIR sensor are all used in the construction of this security system.Telegram is an android app that can be used to communicate data for display on a smartphone [7].A biometric method is one that uses a person's bodily traits or behavior to identify or recognize them [8].One of the most dependable biometric technologies is fingerprint recognition.The technique that has been most frequently used up to this point is fingerprint recognition, while iris recognition and face special features have not been widely used for financial reasons [9].
A microcontroller is a programmable electrical component that can carry out programmed actions.In order to create a whole computer on a single chip, the microcontroller has been fitted with supporting peripherals.To put it simply, the microcontroller is an integrated circuit (IC) made up of counters, clock circuits, parallel I/O, RAM, and ROM [10].A board called Arduino Uno is based on the ATmega328 microcontroller.This board contains a USB connection, a reset button, a power jack, six analog inputs, a 16 MHz crystal oscillator, and 14 digital input/output pins [11].An antenna and PCB board based on the ESP32 processor are used in the WiFi and Bluetooth dual mode development board known as the ESP32 cam.Due to its applicability for smart home appliances, wireless control and monitoring, wireless identification, and other IoT applications that wireless or Bluetooth networks, the ESP32 cam is typically used for a variety of IoT applications [12].
An electronic gadget called a fingerprint sensor is used to take a digital picture of a fingerprint pattern.A direct scan from digital processing produced the image, which was then saved in a database storage memory as a picture of the fingerprint surface that was utilized for matching.The fingerprint scanner system may take a user's fingerprint image and determine whether the flow pattern of the fingerprint in the captured image matches the flow pattern in the database.Then, using serial communication, this sensor transmits fingerprint ID data [13]- [14].An infrared-based sensor is called a PIR (Passive Infrared Receiver).Contrary to most infrared sensors, which are made up of a phototransistor and an IR LED Unlike IR LEDs, PIR does not emit anything [15].The sensor only reacts to energy from passive infrared rays, as the name "Passive" suggests, which are emitted by every object it detects.The human body is often an object that this sensor can detect [16].The KY-024, a linear magnetic module with a hall sensor on it, is the magnetic sensor utilized in this project.This particular kind of sensor can identify the existence of a magnetic field and deliver a voltage proportionate to its strength and direction [17].

Research Methods
Engineering research is what this study falls under.Engineering research is a non-routine design activity, so it produces new contributions in the form of products and prototypes as well as processes and methods.If a design activity refers to specific design standards or codes, it is not an engineering research activity.Engineering research activities discuss design activities that incorporate relatively novel concepts.The six main steps of engineering research are ideas and task clarity, conceptual design, geometric arrangement and functionality, detailed design, prototype/model creation, and tool testing.The first step is ideas and task clarity, followed by conceptual design, detailed design, prototype/model creation, and tool testing.The study's final findings can be applied to technique and testing improvements as well as enhancements to the design process itself [18].
The physical quantities present in the system are measured as part of the data gathering methods used in this investigation.In this study, data on sensor output voltage, program changes, tool accuracy, and tool precision were used as data gathering strategies.There are three types of sensor voltage characteristic data: magnetic, PIR, and fingerprint sensors.By looking at the fingerprint template that is read on the fingerprint sensor, it is possible to determine the properties of the sensor.When comparing the sensor output voltage with variations in distance, magnetic and PIR sensors' properties can be seen.The tool's accuracy data is derived through a comparison of each sensor's output.to assess the tool's precision using many measurements of the same subject.The final step involves evaluating the entire tool to determine whether the smart home security system is performing as intended.The system's components are placed geometrically in accordance with their respective roles.The arrangement of the Arduino-based smart home security system using biometric recognition can be seen in Figure 1.A power supply, fingerprint sensor, magnetic sensor, PIR sensor, ESP32 camera, Arduino as the system controller, Android, and solenoid serve as the components of the home security system shown in Figure 1.Images taken by the ESP32 camera are displayed using the Telegram app on Android.

Design Specification
The tool's performance specification identifies the sensor component that makes up the system through testing and data analysis to determine whether or not the tool is performing as intended.ESP32 Cam modules, magnetic sensors, fingerprint sensors, pir sensors, and tool mechanics are used to illustrate the tool's performance requirements.The component of sensor on the system is shown in Figure 2. shows the three pins-Vcc, Gnd, and out-that make up the PIR sensor.In order for the sensor to detect motion and function in accordance with the provided program, a PIR sensor circuit is required.Figure 2 (c) shows the circuit for the magnetic sensor.The magnetic sensor contains 4 pins, which are labeled Vcc, Gnd, DO, and AO is shown in Figure 2 (c).The magnetic sensor circuit is required so that the magnetic sensor can recognize the magnet that will subsequently be applied to the door and function in accordance with the predetermined program when the door moves.The ESP32 cam circuit is shown in Figure 3 .

Figure 3. ESP32 cam circuit
The ESP32 cam has 16 pins, including the Vcc pin, three Gnd pins, 5V pins, 3.3V pins, IO0 pins, IO2 pins, IO4 pins, IO12 pins, IO13 pins, IO14 pins, IO15 pins, pins IO16, VOT pins, and VOR pins.The smart home security system requires the ESP32 cam circuit in order to connect to the internet and send pictures to the Telegram app.The next step is to create mechanical tools when all the circuits have been connected.Using plywood and acrylic, a prototype of the smart home security system's mechanical construction will be created.Both acrylic and plywood make good bases and sensor protectors.Figure 4 shows the smart home security system's workings.The mechanical design of the full Arduino-based smart home security system with biometric recognition is shown in Figure 4. Arduino, along with other electronic components, are placed in an acrylic container, and all sensors are connected as per Arduino.

Performance Specification
The design specifications are based on the sensor characteristic data used.The sensors that are employed are magnetic, PIR, and fingerprint sensors.Table 1 displays the fingerprint sensor's characteristics.

Table 1. Fingerprint sensor characteristics
Table 1 shows the data of ten fingerprints from person.The fingerprint template from the sensor output will be compared with the fingerprint template that has been stored on the sensor.Use the SFG demo program to display the fingerprint template in order to see the fingerprint sensor's characteristics.By adjusting the distance between the sensor and the object, the properties of the PIR sensor can be observed from the sensor output voltage.A voltage of 3.3 volts is applied to the input sensor, and the sensor output is presented as high and low data, where high corresponds to 3.3 volts and low to 0 volts.Figure 5 displays the PIR sensor's characteristics.

Num
Registered Fingerprint Template Figure 5 shows that, based on the sensor's characteristics, it is unable to detect things beyond a specific range.The sensor can identify objects at a distance of 1 m to 4 m, but it is unable to do so above that point.By adjusting the distance between the magnet sensor and the magnet, it is possible to determine the parameters of the magnetic sensor from the sensor output voltage.5 volts is the sensor's input, and 5 volts and 0 volts, respectively, are the sensor's output in the form of high and low data.Figure 6 displays the magnetic sensor's characteristics.

Magnetic sensor characteristics
Figure 6 shows that the magnetic sensor's output will be low in the presence of a magnet.The sensor can detect magnets at a distance of 0.4 cm to 1.6 cm; over that distance, the sensor's output voltage increases and it is unable to detect magnetic fields.
The smart home security system's accuracy is determined by comparing the sensor voltage listed on the data sheet with the sensor voltage actually measured on the magnetic and PIR sensors.The fingerprint reading template that is registered with the fingerprint that will be read directly is the data that is compared on the fingerprint sensor.Table 2 displays the fingerprint sensor's accuracy.The accuracy of the fingerprint sensor is displayed in Table 2.The words "the compared fingerprint template is matched" may be seen on each finger.Consequently, the fingerprint sensor is accurate.Table 3 displays the magnetic sensor's accuracy.Table 3 displays the magnetic sensor output voltage accuracy for the smart home security system.Here, the magnetic sensor output voltage accuracy is nearly equal to or close to the actual output voltage, with an accuracy rate of 98.92%.Table 4 displays the PIR sensor's precision.In Table 4 shows the accuracy of the PIR sensor output voltage on the smart home security system.The PIR sensor's accuracy in measuring output voltage is nearly equal to or close to the real output voltage, with an accuracy rate of 98.86%.
The output data of the fingerprint sensor, magnetic sensor, and PIR sensor are used to determine the precision of the smart home security system.Ten times the sensor measurement is used to measure with precision.The output data from each sensor is compared to it.The fingerprint sensor will compare the fingerprint template on each used finger ten times.Following that, the output voltage of the PIR and magnetic sensors will be compared ten times for each used distance.Table 5 provides information about the fingerprint sensor's accuracy.

Experiment
To- According to Table 5's data, every fingerprint reading in every experiment was deemed to match, indicating that the accuracy of reading the fingerprint sensor on the right index finger of an Androidbased smart home security system is very good.Table 6 displays the results of the magnetic sensor's accuracy at a distance of 0.4 cm.According to Table 6's data, the Android-based smart home security system's magnetic sensor measurement precision at a distance of 0.4 cm appears to be very good, with a percentage of precision obtained of 97.87%.Table 7 displays the PIR sensor's precise readings at a distance of 1 m.As can be seen from the data in Table 7, the Android-based smart home security system's PIR sensor measurement precision at a distance of 1 m is very good, with an accuracy rate of 99.56%.

Registered
Running an Android-based smart home security system enables general tool testing.Four participants are needed to test the instrument, but only two of them have their fingerprints recorded so that the security system can recognize them.It will be possible to determine later whether the security system's response was as intended.Table 8 shows the result of the tool test table.The tool appears to function in accordance with the inputted program, according to the results shown in Table 8.The solenoid for the door lock will open and the alarm will stop sounding when a user with a registered fingerprint accesses the device.If a user without a registered fingerprint tries to enter the door, the solenoid will remain closed and the alert will still sound.The tool's operation is in line with what is intended.The results have been acquired in accordance with the research objectives based on the analysis that has been obtained in the form of tables and graphs.The latest results are included in performance specifications and design specifications for biometric recognition-based smart home security systems based on Arduino.The functions and circuits of each sensor are used to determine the device performance specifications, and the accuracy and precision of the data analysis performed on the Arduino-based smart home security system are used to determine the tool design specifications.The characteristics of the fingerprint sensor, PIR sensor, magnetic sensor, accuracy, tool precision, and tool testing are used to derive the design parameters for Arduino-based smart home security systems.The results from the fingerprint sensor show that each finger's template is compared to the fingerprint template that was immediately captured.The accuracy of the PIR sensor is 98.86%, according to the accuracy percentage.Additionally, the magnetic sensor's accuracy is 98.92% accurate.Each fingerprint reading that is compared matches according to the findings of the fingerprint sensor's accuracy test on the right index finger.The PIR sensor has a precision accuracy rate of 99.56% at a distance of 1 m.The magnetic sensor has a precision percentage of 97.87% at a distance of 0.4 cm.During the tool's overall testing, it is checked to ensure if it is functioning in accordance with the provided program.According to the results, the Arduino-based smart home security system operates in accordance with the provided program.

Conclusion
The results of the design specifications of the Arduino-based smart home security system are seen from the operation of the fingerprint sensor circuit for fingerprint reading, the PIR sensor to detect the subject, the magnetic sensor to detect the magnet on the door, and the ESP32 cam module circuit so that the system can connect to the internet and can send images to telegram.When the fingerprint sensor reads the registered fingerprint then the Arduino will make the door lock solenoid open so that the door can open.The results of the performance specifications of the Arduino-based smart home security system are obtained from the characteristics of the fingerprint sensor, PIR sensor, magnetic sensor, accuracy, accuracy of the tool and tool testing.For accuracy on the PIR sensor at a distance of 1 m has a percentage of accuracy of 99.56%.For the accuracy of the magnetic sensor at a distance of 0.4 cm, it has a percentage of accuracy of 97.87%.For the accuracy of the security system work on one subject, the results are very good.

Figure 1 .
Figure 1.Arduino-based smart home security system arrangement

Figure 2 .
Figure 2. The component of sensor system (a) fingerprint sensor (b) PIR sensor (c) Magnetic sensor

Figure 4 .
(a) Front view smart home security system mechanic, and (b) Back view smart home security system mechanic Magnetic Sensor 5th International Conference on Research and Learning of Physics (ICRLP 2022) Journal of Physics: Conference Series 2582 (2023) 012025 IOP Publishing doi:10.1088/1742-6596/2582/1/0120255

Table 2 .
Fingerprint sensor accuracy

Table 3 .
Magnet sensor accuracy

Table 4 .
PIR sensor accuracy

Table 6 .
Magnet sensor precision

Table 8 .
Tool test results