Influence of fixed points
Elapsed time per a click which are observed at the last day by proposal systems and existing devices is shown in Figure 8. The elapsed time of when the elbow was used and the participant seated seated with both hands occupied is 2 seconds slower than that of the elbow where the participant was standing with both hands occupied. This is because the users cannot move the elbow freely when they have something in both hands. The movement range of fingers in one hand is too wide to shorten the pointing time. Furthermore, the factors of slow pointing time in one hand are that a finger movement affects the movement of the other fingers and that distance is far from the target of pointing which is located on the left edge of the screen because pivot points are located at the right edge. On the whole, elapsed time of feet is longer than hand, which is considered to be due to the difficulty of stopping for clicking.
The elapsed time of the proposed method is long than that of existing devices. However, users need to have one hand free to be use the trackball. Thus, users cannot use it when both hands are occupied. To use Wii remote, the users needs to move his/her head. This makes it inconvenient because of the lack of eyesight stability, and the situation where users is in crowds is danger. Moreover, moving the head feels strange. Also, this draws undesired attention to the user. On the other hand, the proposed method can be operated in various situations by switching the operating parts.
Relationship between distance and elapsed time
The relationship between movement distance of cursor and elapsed time observed on the last day is shown in Figure 9. The elapsed time of each position of the display shown in Figure 10. The darker the color, the longer the elapsed time. We confirmed that the most of the elapsed time is not the transferring time but the click time because elapsed time increased very little even if the distance from the icon was far from the cursor. When the target located at the screen edge, the elapsed time sometimes becomes very longer if the distance to be moved was short (Figure 10(a)). Thus, we need to simplify pointing at the screen edge. The difficulty of pointing at screen edge affects the total elapsed time.
5. MENU STRUCTURE FOR XANGLE
The arrangement of icon and menu structure in a GUI are based on Cartesian coordinates. This structure is quite good for usage in a traditional environment. However, when using the proposed method, control was achieved by the angles of two lines. We proposed a menu structure suited for Xangle. The menu items lie in both the lines shown in Figure 11, and the items can be selected by moving only one fixed point. The following steps explain how to point at menu items First, the menu items show after pointing at a particular position. Second, the menu items lie in one of the lines. Third, the sub-menu items lie in another line after one of the menu items is selected. Finally, we can select the submenu items.
We evaluated the elapsed time for when the upward finger motion and shin were used. We used two menu structures, the proposed and the traditional are. There were some sub-menu items. There were five menu items and four sub-menu items of each menu. In evaluation, we measured the elapsed time which is for when participants click a sub-menu items 20 times. The target item is selected at random. The method of clicking is equal to the method in Section 4. The participants were the same in Section 4. In the evaluation, participants wore an HMD with an 800 x 600 pixel screen.
The result is shown in Figure 12. It shows that elapsed time with the proposed menu structure is shorter than the elapsed time with a traditional menu structure. Therefore, we confirmed that Xangle can be used to shorten the elapsed time by rearranging menu items. The effect of the proposed menu structure when using the shin is better than with the upward finger. The proposed menu structure simplified the operation because the menu item can be selected by operating only one axis. We consider that this decreases the shaking of feet.
6. ACCESSIBILITY FEATURE
From the result of Section 4, it became apparent that the shake of fixed position was the factor of decline in the pointing speed. Thus, we propose two accessibility features below.
Brake function
This uses an accelerometer which is not used for pointing, which decreases shake by switching the speed of cursor. The movement speed of cursor is limited by the accelerometer. Specifically, there are 3 types, Normal, Slow, and Stop. When the type is Slow, user can know the type because the lines changes from black to red. When the type is changed to stop, click is done at a time.
One axis stop function
This uses an accelerometer which is not used for pointing, which decreases shake by stopping the axis which achieved the target on the assumption that the factor of shake when it operates