Student Research Circle of Biometrics, Image and Signal Analysis and Processing at the Faculty of Computer Science of Bialystok University of Technology

Students in the Research Circle of Biometrics, Image and Signal Analysis and Processing can recognise an active Internet user by the way they type and use a computer mouse. Together with the Medical University of Bialystok, they want to develop a system that can monitor the state of the eye itself.

– Informally, the Circle was founded in August 2015. – recalls Maciej Szymkowski, MSc, Eng, assistant in the Department of Digital Media and Computer Graphics of the Faculty of Computer Science of Bialystok University of Technology. – My colleagues and I were inspired by a series of lectures of Professor Khalid Saeed.

And so, with the start of the winter semester, students persuaded Professor Khalid Saeed, Head of the Department of Digital Media and Computer Graphics at the Faculty of Computer Science of Bialystok University of Technology, to found the circle and become the supervisor to the young biometrics enthusiasts.

– The professor inspired us, suggested which samples and signals we should handle and what we should process. – recalls Maciej Szymkowski, MSc, Eng.

Biometrics is aimed at methods of automatic recognition of people based on their physical characteristics. An example of a biometric measurement device using which specific people can be identified is the iris recognition system, which registers the image of the iris.

– In the circle, we deal with biometric, physiological and behavioural features – explains prof. Khalid Saeed, the circle supervisor. – For example: how to recognise a person through fingerprints, iris, face and others. The way people walk is an example of a behavioural feature. A person can also be recognised through handwriting. The students themselves wanted to develop in this direction and that’s why we founded the circle. Many people joined right away.

During classes in the biometrics circle, its members learned about algorithms, methodology, image processing and analysis. Biometric features were examples of such images. The circle grew and BUT started to purchase the equipment needed for research, such as fingerprint readers.

Thanks to prof. Saeed’s contacts with the AGH University of Science and Technology in Kraków and Motorola Solutions in Kraków, one of the leading technology centres of Motorola Solutions – the American telecommunications giant – in the world, Bialystok University of Technology obtained two grants allowing it to build a biometrics laboratory.

– We bought cameras for face, iris and retina registration. – Prof. Saeed enumerates.

Students quickly began to compile valuable summaries of their activities within the circle.

– The first and very interesting paper was published as early as 2016. – recalls Maciej Szymowski, MSc, Eng. – It was about typing speed and mouse movement – combination of the two. One could wonder: what does mouse movement have to do with determining biometric parameters? Everyone moves the mouse the same way. Surprisingly, it turned out that everyone does it a little differently. The movement and motorics are different. Even button clicking is different.

– Thanks to the cooperation with prof. Saeed, we established cooperation with the Medical University of Bialystok. – rejoices Maciej Szymowski, MSc, Eng. – We worked jointly on systems for diagnosis of retinal diseases, primarily of diabetic retinopathy, which we summarised in the form of a publication as well.

The students worked not only on colour images of the retina, but also on OCT images.

– I like to call them “eye tomography”. – explains Maciej Szymowski, MSc, Eng. – We researched exudative lesions. In our Department of Digital Media and Computer Graphics, we try to develop a complete system of examination based on the research results together with phone software enabling patients to monitor the state of their eye by themselves instead of going to a doctor.

Back in the circle, its students created custom multimodal biometrics systems.

– We’re still working on the system involving three features taken from the finger. – reveals Maciej Szymowski, MSc, Eng. – The work on the software and hardware part is currently ongoing and will result in a complete device.

Although pandemics hindered the work of biometrics enthusiasts, they did not give up.

– Until now circle meetings were held online and we focused mostly on presenting a short lecture on a selected issue at the beginning, and then on the live coding part concerning biometrics itself and the part concerning .NET platform. – explains Aleksandra Kasperuk, President of the circle. – We brought up, for example, binarization algorithms and various kinds of image processing. Image processing can consist in applying specific filters to an image, de-noising or binarisation – for example image skeletonisation.

From such an image, characteristic points are read out, coordinates are recorded and it all makes its way to the database. Information can then be extracted from the database and compared with a sample from a specific person.

– There are more and more students willing to join us, because we teach them something new. Sometimes these things can only be learnt during lectures, but in the circle there is also practice involved. Some of the students stay at the University and pursue further education. – rejoices prof. Khalid Saeed. This is what happened to Maciej Szymowski, the first President of the board of the Student Research Circle of Biometrics, Image and Signal Analysis and Processing.

– Back in the circle, we could attend both national and international conferences. – recalls Maciej Szymowski, MSc, Eng. – In 2016, for example, we attended the CISIM conference in Vilnius. In 2017, we participated online in a conference held in India. In the same year we also had the opportunity to attend a conference in Japan. Thanks to the Professor, we could also attend events such as the Data Science Summit or What the Hack conferences on cyber security. Although the Data Science Summit conference is more about data processing, we were able to establish cooperation with various institutions that were present there.

Some students are interested in scientific development, while others see biometrics as a passion or maybe even an opportunity to get an interesting job. After all, nowadays biometrics is used mainly as a means of control of access to protected spaces or authorising users to access certain data, devices or software. Its primary function is to prevent unauthorised access to cash machines, personal computers, computer networks, mobile phones, home alarm systems, door locks, processor cards and others. Moreover, in public facilities and companies, biometric systems help in the search for whereabouts of specific persons and recording working time.

– Benefits of circle membership include primarily the development of one’s interests – emphasises Aleksandra Kasperuk. – But it’s also just getting together or meeting on Discord or in the lab in person. You get to meet people interested in similar things and this may lead to some cool projects. We always look forward to seeing new members regardless of their current semester. There’s something for everyone.

Student Research Circle of Biometrics, Image and Signal Analysis and Processing

The scope of scientific interests of the circle:

• Digital image processing and analysis
• Signal processing and analysis
• Electronic system analysis and modelling
• Biometrics
• Pattern recognition
• Internet of Things
• Security of applications, devices and computer networks based on biometric features