By using deep learning for small object detection,a novel detection method for key components of elevator landing doors is proposed,which can detect,analyze,and judge the abnormal state of key components of elevator landing doors under normal operation of elevators.In order to verify the detection and analysis equipment for key components of elevator landing doors designed in this article,an experimental platform for detecting and analyzing key components of elevator landing doors has been established.Through on-site verification,the system has achieved functions such as detecting and analyzing key components of elevator floor doors in use.