Background
In the past decades scientific understanding of the impact of light on, for instance, wellbeing, performance, circadian rhythms and sleep has greatly increased. It was for instance found that several disorders, like sleep disorders and seasonal effective disorder, can effectively be treated using light. The NWO project OptiLight aims at integration of human-optimized light control in existing lighting infrastructure. I.e. the lighting in a room will ideally automatically be adjusted to meet the requirements of one or more persons in that room.
Clinical studies have shown that eye light exposure is a main driving force for the day-night cycle of animals, called the circadian cycle. Vice-versa, it has also been determined the sensitivity to light varies over the cycle. Therefore, determining the phase in a circadian cycle of a person is important for determining an optimal control strategy for lighting conditions in a room. However, it is currently not possible to directly measure the circadian phase in ambulatory conditions because it is determined by a large number of cells located in the suprachiasmatic nucleus, a part of the hypothalamus. Therefore, instead more easily measurable indicators that are modulated by the circadian phase, like body temperature and heart rhythm variability, are measured and the circadian phase is estimated by analyzing their modulation pattern.
Since light exposure is the main driving force for the circadian cycle, a person’s circadian phase can also be estimated by analyzing his/her eye light exposure. Currently, light exposure is determined using wearable devices. But each of these wearable devices has its downside: wrist-worn devices tend to be covered by a sleeve, head worn-devices block view or are too heavy, etcetera. Furthermore, such a wearable device is not connected to the existing lighting infrastructure by default and one could think of a number of reasons why it is not desirable to connect a personal device to a public infrastructure. Therefore, it is preferred to create a systems in which the existing lighting infrastructure is able to detect persons in a room, their light exposure, etc. and determine each person’s circadian phase. This way such a system can determine the optimal lighting situation for a person independently.
Description
The project aims to improve the estimation of eye light exposure of persons in a room. With this in mind, the project can be split into a number of phases:
Develop a system that determines the light distribution in a room using one or more IoT device(s) placed at fixed positions in the room.
Using the same IoT devices, detect persons in the room using video/image analysis techniques.
Determine the facing direction of each person using (3D) image analysis of the acquired video data.
Combine the measured data to be able to determining the light exposure at eye level of each person in a room. The system is targeted to be integrated into building lighting infrastructure. This means the system targets are: low cost, low power and low maintenance.
Keywords
Imamage analysis, Video content analysis, Smart Devices, IoT, Human Circadian Rhythm, Lighting, Healthcare
Main supervisor
Jochem Bonarius-de Hoop, MSc (j.h.bonarius-dehoop@tue.nl)
Other involved staff members
Prof. dr. ir. Jean-Paul Linnartz, Thijs Kruisselbrink, MSc., Samantha Peeters, MSc.
Location
Full-time position at TU/e. Cooperation with other faculties: Building Physics and Services, Building Lighting and Social Sciences, Human-Technology Interaction.
Elective courses
Suggested (but others are possible):
5XSF0: Fundamentals of signal & video analysis OR 5XSA0: Introduction to medical image processing
5LSE0: Multimedia video coding and architectures
5LSH0: Advanced video content analysis & video compression