基於影像策略之前導及追蹤輪型機器人之軌跡追蹤控制

在本篇論文裡,我們提出一個轉換策略來轉換輪型機器人為前導輪型機器人或追蹤輪型機器人。我們使用兩部輪型機器人,分別為X80及i90機器人。當輪型機器人以自身位置當做座標中心點,以單眼攝影機取得即時影像,做灰階化處理,經過梯度及建構影像金字塔的處理來進行動態參數的估測。在得到參數後,我們以連續影像相減法將移動物分割出來成為前景,並且以影像區塊法算出移動物重心,並且該輪型機器人轉為追蹤輪型機器人,若無法取得則成為前導輪型機器人。成為前導輪型機器人後,會以紅外線偵測前方是否有障礙物,若無則繼續前導,若有則以影像二值化將障礙物分割成為前景,並且以影像區塊法算出移動物重心後進行橢圓避障後繼續前導。最後不管前導或追蹤輪型機器人皆會判斷是否到達最後目的地,如果沒有則會重複上述動作,直到到達目的地。最後本論文將以實作來驗證我們所提出方法的準確性與有效性。

關鍵字

避障；動態參數；影像處理；輪型機器人

並列摘要

In this thesis, a switching control strategy is proposed for mobile robots to serve as a guiding or following one. Two mobile robots, X80 and i90, are used for experiments. Each mobile robot is setup with a single eye camera to capture the image of the guiding mobile robot or obstacle. The image is turned into a gray form, and then to the gradient for each point of each level of Gaussian pyramid to estimate the motion parameters (disturbance) between two robots. Two successive images are used to catch the moving object to estimate or recognize the foreground. By calculating the gravity of the moving object, the following mobile robot will follow the guiding one. On the contrary, the following mobile robot turns to serve as a guiding mobile robot when there is no gravity to follow. The infrared sensor is used to sense whether the obstacle exists on the guiding trajectory or not. If there is an obstacle, we will use image binarization of red color to get the gravity of the obstacle, and do the collision-free path. By checking arriving the destination or not, the mobile robot will be stopped or keep going until the final destination is achieved. Finally, the experimental results will be used to show the effectiveness of the proposed method.

並列關鍵字

collision-free path ； motion parameter ； image processing ； mobile robot

參考文獻

[18] M.-Y. Shih, C.-H. Huang, J.-H. Kao, and Y.-T. Wu, “Real-time Hybrid Moving Object Detection on Moving Platforms, ” ICL Technical journal, no. 124, pp. 63-70, 2008.

[1] C. C. Chen, “The Design of Autonomous Vehicle with Real-time Object tracking,” Ph.D. dissertation, Dept. Elect. Eng., National Central University, Taoyuan, Taiwan, 2007.

[2] R. P. Chen, “The Research of Distance Estimation for Moving Car-Like Robots Based on Real-Time Image Recognition,” Ph.D. dissertation, Dept. Elect. Eng., Tatung University, Taipei, Taiwan, 2010.

[4] H. Caner, H.S. Gecim, and A.Z. Alkar, “Efficient Embedded Neural-Network-Based License Plate Recognition System, ” IEEE Transactions on Vehicular Technology, vol. 57, no. 5, pp. 2675-2683, September 2008.

[5] H. Sheng, C. Lee, Q. Wen, and Z. Xiong, “Real-Time Anti-interference Location of Vehicle License Plates Using High-Definition Video, ” IEEE Intelligent Transportation Systems Magazine, vol. 1, no. 4, pp. 17-23, 2009.

國際替代計量

基於影像策略之前導及追蹤輪型機器人之軌跡追蹤控制

未授權

主題瀏覽