The portfolio of solutions for retail and smart cities of A.I. Tech is enriched by new exciting functionalities.
Indeed, it is possible to extract additional data from people detected by the video analytics applications. In particular, thanks to the combination of the application AI-CROWD-DEEP with the new video analytics capabilities integrated in AI-DASH it is possible, through t body analysis, to identify the gender of a person, clothes’ color (distinguishing from upper and lower body part) and the presence of accessories like bag and hat.
The extraction of this information is of particular interest in both the retail and smart city areas. In fact, in the first case, it allows to characterize the customers who access the shop; while, in the smart city context it allows to improve forensic search operations on people, thanks to the possibility, for instance, to automatically identify all the women who stopped for more than 5 minutes in the station, wearing a red jacket, hat and bag.
Furthermore, in addition to people counting, AI-CROWD-DEEP allows to characterize the behavior of individuals, evaluating the dwell time (i.e. the time spent by people in specific areas of interest) or identifying anomalous behaviors, such as stationary people, people moving in an area for a long time (loitering). Finally, it allows the recognition of abnormal behaviors of groups, such as gatherings or overcrowding.