基於IMU人形機器人行走之影像穩定

本論文主要為設計一個在人形機器人行走的過程中影像穩定的方法，透過回授的感測器數值與馬達角度推測攝影機的位置與視線的方向，再予以影像的資訊適當的補償。人形機器人在行走時，為了行走的穩定性，會有一定幅度的身體搖擺以將身體的重心移動至走路的支撐腳上，而這個身體的擺動卻影響了影像辨識，導致在計算目標的位置時產生誤差，因此本論文設計一個影像穩定系統，利用安裝在人形機器人身上的IMU感測器偵測人形機器人的姿態並校正影像資訊。在IMU中，通過陀螺儀偵測旋轉的角速度，利用FPGA系統負責積分成旋轉角度，加上加速度計所估測的角度，在FPGA內嵌軟核心NIOS中使用卡爾曼濾波估測身體傾斜度。IMU偵測的人形機器人姿態搭配運動學求得頭部攝影機的姿態，在IPC系統中計算攝影機的位置偏移與視線角度的改變，並對於影像資訊作適當的校正。

關鍵字

人形機器人；影像穩定；姿態估測；慣性量測儀；運動學

並列摘要

This thesis presents a method for image stabilization when humanoid robot walking. Sensor feedback and motor angle situation are used to estimate the position and sight direction of camera of humanoid robot to make up for image information. In order to walk stably, humanoid robot has to swing body and move the center of mass of body to supporting leg. The swing of body effects image recognition to cause some errors for image object localization. This thesis designs an image stabilization system that getting sensor value from Inertial Measure Unit (IMU) for estimating robot attitude to correct image data. In IMU, a gyroscope is used to sense the angular velocity turn into the rotation angle by integration which is calculate in FPGA system. The rotation angle will adjust by Kalman filter in NIOS with the other angle estimation result from an accelerometer. IPC system determines the camera position and direction of sight according to the robot attitude from IMU and camera attitude from kinematics. Then IPC system correct the image information.

並列關鍵字

Humanoid Robots ； Image Stabilization ； Attitude Estimation ； Inertial Measure Unit (IMU) ； Kinematics

參考文獻

[17] 周民偉，大型人形機器人雙足行走步態之設計與實現，淡江大學電機工程學系碩士論文(指導教授：鄭吉泰)，2014年。

[4] F. Lange, J. Werner, J.Scharrer, and G.Hirzinger, “Assembling Wheels to Continuously Conveyed Car Bodies Using a Standard Industrial Robot,”IEEE International Conference on Robotics and Automation, pp.3863-3869, May. 2010.

[5] Z. Yang, K. Ito, K.Saijo,K.Hirotsune, andA.Gofuku, “A Combined Navigation Strategy by a Steering Wheel and a Mouse for a Tank Rescue Robot,”IEEE International Conference on Robotics and Biomimetics, pp.239-244, Aug. 2004.

[7] K.Mikami, and K. Ohnishi, “A Method of Adapting Motion to Depressed Environment for Biped Robot,”IEEE International Workshop on Advanced Motion Control, pp.595-600, Mar. 2010.

[8] DARPA

國際替代計量

基於IMU人形機器人行走之影像穩定

全文下載

主題瀏覽