Analysis And Design Of Helmet Pose Measurement System Based On Inertia And Vision Combination

In modern weaponry,helmets are not only a protective device,but also a powerful information platform.In order to solve the increasingly complicated problem of system operation such as modern aircraft and on-board radar,modern aircraft generally adopt a helmet display with a helmet pose measurement system.One of the key technologies of the helmet pose measurement system is to accurately and quickly measure the relative position of the helmet and the nacelle,and deliver the appropriate image to the helmet display so that the pilot and the onboard equipment can accurately perceive the battlefield environment.From the perspective of pilot helmet equipment in other countries,although there are electromagnetic,electromechanical,visual and other helmet attitude measurement systems,there are widespread shortcomings such as overlapping with infrared spectrum,slow update rate,poor reliability,and low precision,which cannot meet the requirements of modern combat aircraft.To solve the above problems,the combination of inertial measurement unit(IMU)and visual tag AprilTag are used to measure the pose of the helmet in this thesis.By establishing the kinematics model and by using the tightly coupled extended Kalman filter model to fuse the two parts of the pose information together,a new method for corner update of the entire filter is proposed.Firstly,the initialization problems of the whole helmet pose measurement system are studied and solved,including the modeling and calibration of the camera,the establishment of the IMU noise model,the definition of the coordinate system involved in the whole system,and external reference calibration of IMU and camera space pose.In addition,the composition of the visual tag AprilTag system and the algorithms involved in the whole system are introduced in detail.Secondly,a pose measurement algorithm based on Extended Kalman Filter based visual tag AprilTag and IMU information fusion is proposed.Through the integration of the two information,the pose measurement output frequency and accuracy of the whole system are improved to meet the high mobility requirements of the fighter.The specific process is as follows:firstly,the overall design of the whole helmet pose measurement system is carried out.The whole system involves two parts of the camera and the IMU.For the selection of hardware,it is required to meet the requirements of small size and light weight,and design a compact,integrated three-dimensional structure to assemble the two parts of the sensor.Then according to the hardware structure characteristics,the error of the fusion of the two parts of the sensor is analyzed,and the system equations of the camera and IMU tightly coupled kinematics model are established.During the motion,it is compensated reasonably,and the nonlinear system equation is converted into linear.Finally,using the visual tag AprilTag to obtain the corner information as the reference point,a new corner detection method is used to update the entire filtering system in real time,thereby correcting the result of the IMU solution.In order to verify the robustness and feasibility of the helmet pose measurement system based on multi-sensor fusion,six group of comparative experiments are designed.The difference between the spatial pose measured by the helmet pose measurement system and the reference value measured by the Optitrack sensor reflects the accuracy of the system.The experimental results show that the pose measurement system has the advantages of strong autonomy,strong anti-interference ability,high measurement efficiency and no influence on the head movement of the pilot compared with the traditional electromechanical and electromagnetic methods.Compared with the image processing scheme,it has the advantages of smaller changes to the aircraft itself and lower cost.