SMART CAR DOOR LOCK SYSTEMS USING DEEP LEARNING

This study investigated the performance of the smart car door lock system using deep learning. In the era of increasing automation, the advancement of science and technology has made people’s lives more convenient, but it must be convenient and safe at the same time. The method used in this study combined facial recognition and fingerprint recognition as the verification method of the car door lock system. Also the research method is mainly divided into three parts: the application of YOLO face recognition,fingerprint recognition, and the integration of the two recognition methods, which are applied to the car door lock system. The face recognition part used the YOLO convolutional neural network, and one thousand facial images of the three people were used to train the recognition system. The weights were obtained after tens of thousands of training. After adjusting the various parameters, facial recognition could be operated. Fingerprint recognition used an optical fingerprint sensor to store the fingerprints of three people in the memory, and used software to control the signal output of the fingerprint recognition. Lastly, the facial recognition signal and fingerprint recognition signal were integrated with the software. The purpose is to achieve that the car door lock will be opened when the two verification methods are correct at the same time. On the contrary, when any verification method is wrong, the car door lock will still be closed