Ковариант


Vision is the arguably most important of our senses. Much of our knowledge of the external world is gathered by using our eyes. People make sense of the world through these built-in sensual cameras.. Eye tracking gives us the ability to judge the physical appearance and visible actions of other people. It also allows us to develop the further perceptions and interpretations that accrue from eye contact, which in turn produces meaningful insights about behavior and performance.Moreover, like detectives, we can observe what other people look at when they are not looking at us.

By knowing what a person looks at we can better understand that person’s behavior. Retailers and FMCG companies use eye tracking to understand how shoppers interact with products in a real shopping environment.

By knowing how – that is, the manner in which a person visually reacts to various stimuli, we can understand and fix some medical problems. Eye tracking is used in various psychology and neuroscience fields to understand how and why eye movements are made and how we gather information visually. Infant and child researchers use eye tracking to study perceptual, cognitive, and social emotional development from birth through early adulthood. Eye tracking in combination with conventional research methods or other biometric sensors can even be helpful for diagnosing diseases such as Attention Deficit Hyperactivity Disorder (ADHD), Autism Spectrum Disorder (ASD), Obsessive Compulsive Disorder (OCD), Schizophrenia, Parkinson’s and Alzheimer’s disease.

Also tracking is used in automotive and aviation to analyze the eye movement of driver or pilot in a difficult situation, visual attention e.t.c.

Currently, the most common methods of eye tracking are eye-attached tracking and optical tracking. Through image processing algorithms, a computer-aided vision system is constructed that isolates the human eye and determines the orientation of the eyeball in the set of primitives of the image being processed.

These systems are quite effective and have found wide application in modern technology, but have a significant drawback - the bulkiness of the device, which most often represents a helmet or a desktop camera installed in front of a person. Also these systems need to have a hard and expensive intellectual block to video recording and image processing. Thus, the device becomes unusable for permanent wearing, which is especially important for medical applications.

The way out of the situation turns out to depend on electric potential measurement systems based on highly sensitive sensors of the electromagnetic field. The main idea is to catch the electrical signals coming into the eye muscles. Human muscles are controlled by the nervous system by means of electrical pulses of variable polarity, which in turn leads to the emergence of an alternating electromagnetic field in space. The engineering task is to catch and filter these signals against the background of the numerous disturbances present in the surrounding space. The second stage is software, which will match the detected signals of the eye muscles and the real orientation of the eyeballs caused by their contractions. Further information on the direction of vision is transferred to the upper level system and used for therapeutic purposes.

To register such weak fields as those induced by eye muscle signals, we used highly sensitive sensors from Plessey Semiconductors.

We applied a typical sensor switching scheme recommended by the manufacturer, tuning our solution with feedback and filters that cut off the first part of the constant noise at the input of the device.

After that we collected several sensors in the sensor field and placed them on a medical mask, providing the service electronics, an independent power source and a channel for transferring data to the computer via Bluetooth.

The “software” part of the project consists of four different modules:

  1. - Sensor microcontroller firmware
  2. - Additional utility software to calibrate the system and select the necessary parameters of signal filtering and processing
  3. - Server software that collects the data and processes the signal “on the fly” to improve the performance of the mobile application
  4. - User application running on different mobile platforms

Sensor microcontroller firmware
The signal from the detector is rather noisy due to the electromagnetic fields around us. We had to choose the high-precision ADC (analog-to-digital converter) and advanced microcontroller to achieve the preliminary signal processing before sending the data through Bluetooth.

Utility software
The main aim of the utility software is to help us determine the best parameters of the systems, such as: the position of the detector on the face, the sizes of the electrodes, the best filtering algorithm to minimize the data transfer rate, etc.

At this point it should be mentioned separately that we also encountered the task of finding the correlation between the specific muscles activity and the data collected from the signal. The first results have convinced us that the problem can definitely be solved. But if we want to achieve real-time operation mode and high precision, the machine learning algorithms must be used.

Server software
The data collected from the devices are processed on the server side since the mathematical models are too complicated for the microcontroller or the mobile processor. Also we can use collected medical information in ophthalmological research. In this case our software might well arrive at a proposal in the form of a medical recommendation concerning the eye's health.

Mobile application
This application provides communication with the sensor device through Bluetooth and arranges the data transfer to the main server through the secured channel. The results of the filtering and processing then are transferred back to the mobile application. The user can have a look at the analytics and browse through the history of the previous measurements. Nowadays lots of people work with the computer the majority of time and have some issues with their eyes as the consequence. This users can see recommendations, alerts and have an ophthalmological consultation.

Контакты:
Email.: office@covariant.info

This email address is being protected from spambots. You need JavaScript enabled to view it.

Напишите нам с сайта:  задать вопрос