Human Action Recognition-Based IoT Services for Emergency Response Management

Emergency incidents can appear anytime and any place, which makes it very challenging for emergency medical services practitioners to predict the location time of such emergencies. The dynamic nature appearance cause delays in services, sometimes lead vital injury complications or even death, some cases. delay may occur as a result call that was made too late because no one present make call. With emergence smart cities promising technologies, Internet Things (IoT) computer vision techniques, issues be tackled. This article proposes human action recognition-based IoT architecture response management. In particular, exploits devices (e.g., surveillance cameras) are distributed public areas detect incidents, request nearest send information. Moreover, this an detection model, based on recognition object tracking, using image processing classifying collected images, modeling. primary notion proposed model is classify activity, whether incident other daily activities, Convolutional Neural Network (CNN) Support Vector Machine (SVM). To demonstrate feasibility several experiments were conducted UR fall dataset, consists activities footage. results promising, with scoring 0.99, 0.97, 0.98 terms sensitivity, specificity, precision, accuracy, respectively.