Pose Measurement Technology With Inertial And Visual Fusion

Pose measurement technology is to obtain object pose information using one or more sensors.Quick and accurate pose measurement is widely used in aerospace,manufacture and robot fields.Inertial pose measurement,using iteration formula,has an advantage in measurement velocity but suffers from slow drift.In contrast,visual pose measurement based on POSIT algorithm is slow but accurate.This paper blends these two methods to be a fusion-measuring method.It overcomes the shortcomings of each method,and retains their respective advantages.The main topics and key techniques are composed of the following three parts:1.The technique of inertial pose measurement,which includes the calibration of coordinate relationships,the iterative algorithm of pose solution,as well as the compensation of angular-velocity drift.2.An inertial and visual fusion pose measurement method using a robust filter.This paper designs an improved H? filter to integrate the datum from above two types of measurements.Then Euler drift error curve can be corrected and updated,according to the difference between the outputs of hybrid measurement and inertial measurement.It is a significant step to adaptively compensate angular velocity drift.The experimental results show that this fusion method can achieve fast-accurate pose measuring in long time.3.A fusion pose measurement method by calibrating inertial measurement results using visual information.Calibration occurs in specific intervals in space.And then the new value of target pose is obtained an improved visual measurement,using global-control-points calibration.