Intelligent agent high-dynamic pose estimation method based on motion coding event plane representation
The invention provides a motion coding event plane representation-based agent high-dynamic pose estimation method, which comprises the following steps of: 1, reading an event stream output by an event camera, and representing each event by a four-dimensional vector; 2, constructing a double linked l...
Saved in:
Main Authors | , |
---|---|
Format | Patent |
Language | Chinese English |
Published |
30.08.2024
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | The invention provides a motion coding event plane representation-based agent high-dynamic pose estimation method, which comprises the following steps of: 1, reading an event stream output by an event camera, and representing each event by a four-dimensional vector; 2, constructing a double linked list, storing pixel triggering time and sequence in a local neighborhood, and carrying out stack updating through an asynchronous event driving thread; 3, constructing event plane representation of motion coding, and realizing environment consistency representation in a high-speed and high-dynamic scene; 4, constructing a semi-dense scene local map in a 3D point cloud form; and 5, realizing real-time estimation of the six-degree-of-freedom pose by adopting a 3D-2D alignment technology on the basis of coding of the motion space-time constraint of the camera based on event plane representation. According to the method, the low delay advantage of the event camera and the natural response to the scene edge are utilized, |
---|---|
Bibliography: | Application Number: CN202410613621 |