| dc.description.abstract | There are vast applications of indoor localization and crowd behaviour analysis. Application domains include hospitals, workstations, malls, supermarkets, festivals, etc. Crowd behaviour analysis becomes very important in the scenario of heavy public gatherings to assist on many use cases such as planning emergency exits, analysing trends, etc. although global positioning systems are very efficient solutions for outdoor positioning and navigation, they fail miserably in an indoor environment. Thus, the importance of building reliable indoor positioning systems comes into the picture. Indoor positioning systems act as a base point for indoor individual behaviour analysis. That in turn fuels the possibility of collective approach to crowd behaviour analysis. Out of all hardware setups, we decided to experiment on wifi signal strength based localization setup since this setup has minimum hardware installation requirement. It is based on the assumption that the majority of the indoor environment has wifi routers already set. We have used one publicly available dataset UJI Indoor localization and another wifi dataset based on our campus, collected by our PhD student. We have implemented both static plots and interactive animations for both user and crowd level analysis points. | |
