A REAL-TIME POLICE DOG ACTION RECOGNITION SYSTEM BASED ON VISION AND IMU SENSORS
Xiaoman Zhan, Qian Huang, Chaozheng Zhu, Guangyun Liu, Xing Li
-
SPS
IEEE Members: $11.00
Non-members: $15.00Length: 07:16
A new human-dog coordination system is in urgent need to enhance the working efficiency of both the police and the police dogs. In view of the problem of sensor data disorder caused by jitters in the state estimation of non rigid animals such as police dog, this demo paper proposes a real-time solution for action reconstruction and video transmission of police dogs based on vision and IMU sensors. These sensors are integrated in a camera helmet and a police dog vest, which have the characteristics of easy wearing, high accuracy and good real-time performance. With the proposed user-friendly GUI, the police are able to work remotely with police dogs with quick reaction and low security risks.