OccuBOT OccuBOT is a python-based web app developed to determine and compare the occupancy of facilities, using security camera frames as its source. It does so by taking advantage of an open source computer vision library (OpenCV), as well as with Flask integrating the front-end with the back. The result is an easy-to-use software designed to improve the daily experiences of its users in both low and high-octane situations.
Abstract Using facial recognition combined with human limb-like recognition, the ML algorithm will be able to determine the number of individuals in a predefined region. Settings can be pre-adjusted for the software to generally predict the approximate scale of the human to be detected. This information can then be used to count the number of people, to ultimately ascertain the occupancy of the room.
How we built it OccuBOT is built with python2.7, using OpenCV for the human recognition functions and Flask to implement the web app propotype, all in under 36 hours by a team of 4 at Hack the North 2017, in Waterloo Ontario.
Practicality The applications of this web app are numerous and wide-reaching; being highly useful in both casual and emergency scenarios. OccuBOT could be used on a daily basis, with simple tasks such as determining which cafe has the shortest lines or which library is the quietest at that exact moment. This eliminates unnecessary commute time between establishments and creates a more relaxed, productive day for the user.
On the rarer but arguably more impactful side, there is the application of OccuBOT in emergency situations during there is no time to waste. If an individual is in a life-threatening state and every second is of the utmost importance, you do not want to show up to an emergency room only to discover that there is a three-hour waiting period. As such, OccuBOT could quite literally save lives by cutting precious time off of wait-times in hospitals.
OccuBOT represents an incredible avenue through which to improve users’ daily experience while also acting as an invaluable tool in time-sensitive situations, all while keeping the anonymity of the user.
Future Applications In addition to OccuBOT’s aforementioned applications, there is also the option of adding further features, owing to OpenCV’s open-source recognition library. The first large feature would be that of demographic testing. By using OccuBOT, not only would the occupancy of the room be determined, but also the demographic of those within it. This would provide inexpensive information to the establishment regarding their clientele. From a commercial perspective, companies would then be able to better appeal and cater to their demographic all while seeing accurate, real-time numbers regarding who is consuming the product/service and when.
Secondly, OccuBOT presents a unique opportunity through which to improve public safety on a large scale. The facial recognition capabilities of this software, while keeping the user anonymous to other users, could be used to detect high-risk individuals and alert the appropriate authorities. This creates an non-invasive means to improve public security.