Improving Vehicle Security using CNN and Internet of Things Based Face Identification

In this modern world, vehicle security remains as a major challenge as the conventional physical vehicle locks are insufficient. This research work aims to explore the potential of integrating face recognition technology for vehicle security. The proposed system introduces a novel deep learning algorithm integrated face recognition approach to authenticate and enable access to vehicle. The main functionality of the proposed system involves training the developed face recognition model with an extensive dataset of facial images, wherein the Convolutional Neural Networks (CNN) are used for performing feature extraction and classification. With this, the proposed deep learning model will accurately detect and authenticate the vehicle user/owner’s face and eliminate the usage of manual vehicle locks. In this way, the proposed approach offers an efficient and robust vehicle security solution aligning with the contemporary technology-driven world. Furthermore, to validate the effectiveness of the proposed system by testing it with real-time human faces. In the real-time implementation, the accuracy of the proposed face recognition system is 87.5%, with a precision of 90%, recall of 85.71%, and an F1-Score of 87.88% in recognizing the vehicle owner’s face and confirm the proposed system as an effective alternative to traditional vehicle security methods.