Real-Time Detection System of Bird Flight Gesture

In the study of bird biological attributes, it is urgent to study its dynamic gesture in long-distance and long-time air flight. However, at present, such research is weak and lack of classification standards, record detection, real-time transmission equipment, background analysis software, etc. Firstly, the bird flight gesture is studied, nine classification and coding schemes are defined, and a flight dynamic gesture detection method based on MEMS nine-axis sensor is designed. The quaternion method is innovatively used to derive a complete set of flight gesture judgment algorithms. The judgment process of left and right direction is given. On this basis, a set of remote real-time detection equipment based on 4G communication is developed, which can transmit the bird flight gesture data and other information back to the background server in real time. At the same time, a background graphical analysis software is developed, which can use the browser to query and display the gesture of multiple flying bird objects in real-time. The experiment shows that the detection and wireless communication time of the system is less than 500 ms per cycle, which meets the requirements of real-time detection.