Smart building management is an essential and promising field, considering the economic and ecological advantages it offers. By actively adjusting lighting, ventilation, and heating parameters, it is possible to reduce a building’s energy consumption without compromising occupants’ comfort. In this thesis, we explored several solutions to estimate the number of people in a room, a crucial piece of information for proactive management. To begin with, we investigated people detection using low-resolution images, evaluating the necessary level of supervision and the effects of implementation in unknown environments. Next, we studied the optimal temporality for recognizing occupants’ actions. Lastly, we delved into integrating temporal context to enhance high-resolution people detection. All of these research efforts were conducted using infrared images to preserve individuals’ anonymity while enabling the widespread application of these methods.
| Date | 27 Nov 2023 |
|---|
| Original language | French |
|---|
| Awarding Institution | - École de technologie supérieure
|
|---|
| Supervisor | Éric Granger (Supervisor) & Marco Pedersoli (Co-supervisor) |
|---|
Dubail, T. (Author),
Granger (Supervisor) &
Pedersoli (Co-supervisor),
27 Nov 2023Student thesis: Master's thesis › Master in Engineering: Automated Manufacturing Engineering