FPGA AND DEEP LEARNING BASED REAL TIME FACE RECOGNITION SYSTEM FOR E VOTING SYSTEM
The deep learning techniques part of the software helps develop error-free and less complex algorithms using a few blocks rather than a thousand lines of code. The data is properly formed as well as trained and processed for correct recognition of the correct vote.
2025-06-28 16:32:40 - Adil Khan
Project Title
FPGA AND DEEP LEARNING BASED REAL TIME FACE RECOGNITION SYSTEM FOR E VOTING SYSTEM
Project Area of Specialization Internet of ThingsProject Summary- This project is a combination of hardware and software that provides the best solution to deal with the "efficiency vs speed" compromise problems, it will be designed to develop the tests and deal effectively with the recognizable system which can be used for several applications, especially for an electronic voting system in Pakistan.
- The deep learning techniques part of the software helps develop error-free and less complex algorithms using a few blocks rather than a thousand lines of code. The data is properly formed as well as trained and processed for correct recognition of the correct vote.
- Significantly, OpenCV and its integration in the premises of the programming software (based on deep learning) also bring many advantages to this project as well as the hardware part ‘ MFRIO 1950 device’ which is in fact the factor of setting highlight of this project which will help to provide an adequate database as well as to process a large amount of data in a real-time environment used for authentication purposes.
- The main objective of this project is totally focused on the electronic voting system, the security of any institution during elections and entering cricket stadiums or other related applications where a delay of a few nanoseconds can cause significant damage. These systems must therefore be implemented, in particular surveillance zones where appropriate control and balance are the main demands to be acquired.
- The proposed designed system is tested in real-time scenarios and can be adapted to the electronic voting system via the NADRA database.
- Implementation of system e-voting and especially for election in Pakistan.
- To design and develop the facial detection and recognition application based on deep learning techniques that reduce the error ratio.
- To recognize the faces with the minimum processing delay.
- To become capable of processing large data in real time environment similar to secure standalone embedded application, without the involvement of personal computers.
- It increase the security by using facial recognition technology specially in surveillance area.
- To reduce the physical interaction.
- Provide an adequate database as well as process large number of data in real time environment used for authentication purpose.
- The implementation of the system proposed with MyRIO. It takes input from the camera by capturing the image and then processes the captured data according to facial recognition algorithm.
- The results of the authorized person have been analyzed via LabVIEW software, when the person in the database as they appear in front of the camera, they will recognize and respond with a green LED as an authorization indicator, in the same way as they will unlock the door in a real-time environment, when the software algorithm is implemented in MyRIO.
- This inclusion makes the process even more efficient and faster, as various built-in facial recognition blocks can be processed in LabVIEW. On the other hand, the hardware part "MyRIO 1950 device" which is in fact the highlighting factor of this project which makes it possible to provide an adequate database as well as to process a large number of data in an environment in real time used for authentication purposes.
- Electronic voting technology aims to speed up the counting of votes, reduce the cost of paying staff to count votes manually, and improve accessibility for voters with disabilities.
- The expenses on election is expected to decrease
- Results can be reported and published faster.
- Voters save time and money by being able to vote regardless of location, which can increase overall voter production.
- The groups of citizens who benefit most from electronic elections are those who live on board, citizens living in rural areas far from polling stations and people with disabilities.
- The electronic voting system can improve election system and reduce ratio of re-election based on miscounting of voting.
- The system is fully based on 4th Industrial revolution-based technologies like loT, Big-Data and Artificial Intelligence, the project technical flow will be under as:
- On physical and software layers various types of sensors, softwares and detectors will be used to detect the face of person, like as face detection, face recognition, real-time environment, NI MyRIO FPGA, LabVIEW, OpenCV integration, NI VISION tool kit, security and electronic voting system.
- The sole purpose is to find the location of the face in the image and extracting its coordinates that are followed by applying the recognition algorithms to match the face from the database.
- The face recognition system is quite flexible than other biometric techniques such as fingerprint, iris scanner, signature, ID card etc.
- It takes the input from the video camera by capturing the image then process the captured data according to algorithms of face recognition. It is responsible for decision making after feature extraction and face detection.
- Compared to traditional computing techniques, this platform of MyRIO gives more optimum and effective results.
| Item Name | Type | No. of Units | Per Unit Cost (in Rs) | Total (in Rs) |
|---|---|---|---|---|
| Total in (Rs) | 79500 | |||
| Camera | Equipment | 1 | 50000 | 50000 |
| Electronic IC | Equipment | 2 | 9500 | 19000 |
| packaging | Miscellaneous | 1 | 10000 | 10000 |
| LED | Equipment | 5 | 100 | 500 |