- 學位論文
應用彈道式移動模型量測三維空間中的瞄準移動
An Application of Ballistic Movement Models for Three-Dimensional Pointing Movements
摘要
在現代日常生活中，人們常需要使用手來執行各種工作，例如以滑鼠控制游標開啟Windows 作業系統中的應用程式。然而，一個手控制移動實際上是經由多個彈道式移動所構成，而人在執行這些彈道式移動的能力直接影響到手控制移動所需花費的時間以及準確度。 本研究目的是以Lin (2010)與Lin et al. (2011)在2D空間中所驗證的彈道式移動模型為基礎，量測在真實3D空間中執行彈道式手部移動所需花費的時間和移動後落點與目標點之間的誤差，並驗證彈道式移動時間模型和彈道式移動變異模型的有效性。 本研究分為兩個實驗，實驗一為測試性實驗，其目的為確定實驗設計的可行性，實驗二為正式實驗，自變項為受測者、慣用手與非慣用手、七個移動距離及七個移動方向，依變項為執行彈道式移動所花費時間與移動後三軸落點誤差變異。 實驗結果顯示：(1)三個自變項(受測者、左右手、距離)對於彈道式移動時間及三軸落點變異皆有顯著影響(p < 0.05)，(2)彈道式移動時間模型成功描述移動時間與距離平方根之間的線性關係(R-S(adj)=99.3%)，(3)彈道式移動變異模型有效描述三軸落點變異和距離平方之間的線性關係，模型分別能對X、Y與Z軸描述84.4%、99.0%及91.3%的資料變異，而三軸誤差變異由高至低依序排列為X軸、Z軸然後為Y軸，(4)左右手在不同移動方向下，由於使用到的手部關節部位不同，所需花費的移動時間與落點誤差變異也會不同，整體而言，移動所使用的關節部位越少，移動時間越快，但在X軸上的落點變異會越大，而Y軸上的落點變異會越小，而因為執行彈道式移動的固定高度因素，所量測到的Z軸落點變異卻是三軸中最大的。 本研究成功量測3D空間中不同移動方向執行的彈道式移動並驗證彈道式移動兩個模型，彈道式移動模型相較於Fitts’ Law (1954)能給予手控制移動在移動速度及準確度上的獨立分析，未來可作為評估輸入控制設備的準則。
並列摘要
In daily life, we use their hands to execute various activities, such as controlling a mouse cursor to open an application program in the windows operating system. However, a hand-control movement is composed of several unit movements, called “ballistic movement”. Hence, the capability (speed and accuracy) for executing ballistic movements directly affects the performance of a hand-control movement. The main purposes of this research were to verified the application of the two ballistic movement times proposed by Lin (2010) and Lin et al. (2011) in a real 3D environment, and to utilize these two models for measuring the ballistic movement time and ballistic movement end-point variability. This research was consisted of two of experiments, in which the first experiment was a pilot study and the second experiment was formal experiment. The independent variables studied in these experiments included participant, dominant and nondominant hand, movement distance and movement direction, and the dependent variables were movement time and movement variability measured in three dimensions. The results showed that (1) three independent variables (participant, hand, distance) had significant effects on ballistic movement time (p < 0.05); (2) the ballistic movement time model successfully predicted the linear relationship between movement time and the distance square root (R-Sq(adj)=99.3%); (3) the ballistic movement variability model successfully predicted the linear relationships between three dimensional end-point variability and the square of distance. The model explained X-axis, Y-axis, and Z-axis errors 91.3%, 84.4%, and 99.0% of data variance, respectively. The order of three dimensional end-point variability arranged from the highest to the lowest was X-axis, Z-axis and then Y-axis; and (4) for both hands, movements that involved less moving joints and segments required shorter movement time and resulted in larger X-axis end-point variability and smaller Y-axis end-point variability. However, we found that Z-axis end-point variability was the largest among the three dimensional measurements. This might due to the consist height of the ballistic movement execution location set in the experiments. In this specific height, participants might maintain difficultly the Y-axis accuracy. This research successfully measured different direction of 3D environment movement performed in different movement directions and verified the two ballistic movement models. Ballistic movement, compared to Fitts’ Law (1954), is expected to analyze movement speed and movement accuracy independently. Future research could utilize the two models for evaluating input control devices.