Image identification technology for door unlock application

• Main Authors: LI, Zi-Rui, 李梓睿
• Other Authors: Wu, Ming-Fang
• Format: Others
• Language: zh-TW
• Published: 2018
• Online Access: http://ndltd.ncl.edu.tw/handle/tm5586

碩士 === 崑山科技大學 === 電機工程研究所 === 106 === When the function of the car is more and more, however the main reason for the most frequent of car accident actually comes from the driver. Most of the safety support systems are based on the detection of external or car body abnormalities and may affect traffic safety as a protection. For the driving behavior of their own will be dangerous and protective system is relatively few, to drive the car door inadvertently caused by the harm prevention system is none. In recent years, the image processing technologies are continuous improved, so the accuracy of face recognition continues to promote. When the hardware in the semiconductor process continues to improve, and the computing speed is more powerful, it is not a difficult task to make the protection system for the detection of driver misconduct. This paper utilizes the excellent computing ability of the Raspberry Pi development board, which is coupled with the Raspberry Pi camera, and uses the advantages of GPIO to control the output signal as a hardware platform. In the process, Open CV image processing is used for the human eye recognition system to confirm that the driver has turned his head and paid attention to the rear of the vehicle and recognizes that the eyes later control the lock in the vehicle door so as to prevent the danger caused by human factors.