Multi-channel Complex for Non-contact Aggressive Movement Detection

The paper presents a multi-channel complex for non-contact movement detection, which consists of bioradar sensors and an RGB-D camera. The proposed system was validated on an experimental dataset of scenarios performed by 10 volunteers (3 males and 7 females) in the age group between 20 and 36 years participating in the study. All subjects gave their informed consent prior to the start of the experiments. During the experiments, the signals from two bioradars were recorded simultaneously with the data from the RGB-D camera. Each subject was asked to perform 11 different types of common everyday motion patterns and aggressive movements, each repeated three times. During the experiments, 330 records corresponding to non-aggressive activities and 30 to aggressive behavior were collected. The proposed classification algorithm showed 98% accuracy and Cohen's kappa of 78% for the non-aggressive/aggressive behavior classification. The future activity will consider expanding the dataset considering different surrounding and lighting conditions.