Fog cloud-assisted IoT-based human identification in construction sites from gait sequences

Human identification on construction sites is critical for minimizing safety mishaps. Existing approaches have shortcomings such as a low recognition rate, workplace locating errors, and alarming latency. These challenges are addressed in this study by developing a Fog Cloud Computing, Internet of Things (IoT)-based Human Identification system based on Gait Sequences. Gait recognition, as a prospective biometric identification approach, has several advantages, including the ability to identify humans at a great distance, without any interaction, and the difficulty of imitating. However, due to the complexity of the external components involved in the collection and sampling of gait data and changes in the clothing style of an individual to be recognized, this recognition technology continues to confront several obstacles in real-time applications. In this study, the purpose is to offer a unique method for gait feature extraction and classification at construction sites. The feature vectors derived from Speeded Up Robust Features (SURF) and Convolutional Neural Networks (CNN) are integrated. The classification is performed by applying a Support Vector Machine (SVM) to increase the recognition rate at the Fog layer. The decision-making, record storage, and monitoring processing are performed in the cloud layer. On comparative analysis, experimental results demonstrate that our proposed model outperforms the existing methods and attained the highest accuracy of 97.19%.