手指復健機之研發

Abstract

根據世界衛生組織統計資料顯示，中風是已開發國家中前三大的死因之一。在我國中風發生率的研究指出，36 歲以上人口的發生率約為千分之三，若以台灣地區36 歲以上人口數接近一千萬，每年的中風新發生數約為三萬人。代表中風後的復健治療需求越來越大，導致復健師與復健載具的需求量也隨之變大。然而復健師數量有限，因此必須藉由大量的復健載具來幫助復健師節省時間與人力成本。復健載具主要分為兩種，一種是復健機針對完全無自我意識的動作功能恢復，另一種則是輔助載具，主要感應患者想要的動作來進行力道的補償。到目前為止許多團隊已開發出多種復健載具，但無法大量製作成產品提供復健師使用主因為體積龐大難以攜帶及製作成本太高不符合經濟效益，因此本研究試著提出新的復健機方便人體使用。 本研究設計一個方便放置於手上的手指復健機，可適用的人體手指長度為2公分至8公分。藉由相機拍攝正常人的手部運動，經由攝影機校正與三維影像重建技術計算出手指關節彎曲的角度，在控制復健機轉動到此角度以便帶動中風患者的手指做相同的動作，如此重複進行來達到復健治療的目的。本研究的控制器採用DE1-Soc其最大優點為平行處理，可以同時處理相機影像、角度計算與馬達控制，來達到即時控制的功能。攝影機校正使用張友正法來得出相機的焦距、光軸偏移量、兩台相機的相對位置關係，並探討與相機不同距離下，其量測點之誤差。 實驗結果顯示，復健機可移動的範圍為78度，手指伸展可達8.4度，而彎曲可達69.56度，而轉速可達3.22RPM，另外可承受的最大扭矩為3kgw*cm。在相機影像的三維重建方面，若能將量測物移動方向與光軸越接近，其誤差會與理想誤差更接近。

According to WHO statistics, stroke is one of the top three causes of death in developed countries. Research of stroke incidence in Taiwan indicates that the incidence of stroke is above three thousandths among the population over 36years of age. If Taiwan's population over the age of 36 close to ten million, the number of new stroke about 30,000 people per year. It means the demand of rehabilitation treatment after stroke is increasing, and the demand of therapists and rehabilitation skeletons will become lager, too. However, the quantity of therapists is limited, so it needs more and more rehabilitation skeletons to save time and manpower. The rehabilitation skeletons divided in two kinds, one is rehabilitation machines which focus on the action function recovery without self-awareness. The other is assistant machines which compensate the force to finish the motions that patient want to do. Up to now, many research teams have developed many kinds of rehabilitation skeleton. But the cause that it can’t make into large number is the volume of their machines is so large that it can’t be easy to carry and the costs are too high to make them. So, this research proposed a new rehabilitation machine for patient easy to use. The research designed a rehabilitation machine for fingers that are placed on human hand conveniently. It can be suitable for the finger which length is 2cm to 8cm. According to the images of normal human hand’s motion that captured by two different cameras, we can calculate the angle of finger by the technology of camera calibration and three dimensions reconstruction. Base on the angle that we calculated, the rehabilitation machine can bring human finger to do the same motion that normal human motion. Doing the same procedures over and over again can finish the physical therapy treatment. The controller of this research is DE1-Soc which can process the images, angle calculation and control motors in parallel, so it can execute in real time. The method of camera calibration is Zhenyou Zhang calibration which can obtain the focal length of the camera, the offset of the optical axis and the position relationship between the two cameras. Finally, we discussed the error of points at different distances from the cameras. The experimental results show that the range of the machine motion is 78 degrees. The extension of finger can reach 8.4degrees and the flexion of finger can reach 69.56 degrees. Then, the maximum of angular velocity is 3.22 RPM, and the maximum torque that can endure is 3 kgw*cm. In the three dimension reconstruction, if the things that we measure move along the optical axis, the error will become small and be closed to the ideal error.