運用多重行動裝置於3D 影像之視覺化技術: 以3D 醫療影像視覺化為例

Abstract

近年來，由於行動裝置(例如智慧型手機)的活動性以及強大的功能，它已成為我們日常生活中不可或缺的必要配備。加上硬體的發展迅速，現今的行動裝置在某些功能上已經可取代個人電腦，有時甚至可比個人電腦做得更多，因此，越來越多有關行動裝置上的3D視覺化技術以及互動技術的研究出現了。考慮到3D醫療影像視覺化的需求，我們提出了一種新式的互動模式，使用者可透過行動裝置與3D醫療影像互動。根據此互動模式我們提出了一個系統multi-plane 3D display system，使用者可以透過多個行動裝置看到3D影像並與3D影像互動。此論文中所提到的3D影像並不是傳統的3D物件，而是一個不同的概念，是將多個連續的2D影像堆疊形成3D影像。在此系統中，使用者只需要轉動行動裝置，就可以看到任何方位角度的3D影像切面，並且可藉由多個裝置的合作，選取想看的切面位置。多個裝置的合作也帶來一個特性 ─ cascaded view，各裝置顯示的切面都是有關聯的。另外，我們也提供使用者其他的方法去改變切面的位置，例如用觸控手勢。加上考慮到使用者經驗，我們提供回朔和註記的功能，讓使用者可以回到先前所看過的特定切面狀態，以及在切面上做註記。目前我們已使用電腦斷層掃瞄影像為3D影像來源實作出此系統的原型。

In recent years, mobile device such as smart phone or tablet PC has become an indispensable device in our daily life not only because of its mobility but also its powerful functionality. Along with rapid development of hardware, in some condition, mobile device can do as much as a personal computer or more. As a result, researches about 3D visualization and interaction on mobile device arise. Considering the demand for 3D medical image viewing, we provide a novel interaction pattern to interact with 3D image using mobile devices. We develop a multi-plane 3D display system which allows user to see and interact with 3D image via multiple mobile devices. 3D image is not a traditional 3D object; it is constructed by stacking continuous 2D image data. In our system, user can see cutting faces of 3D image in any orientation by rotating mobile device. The rotation and orientation are detected by accelerometer and magnetic sensor. user can also select the position of cutting face in 3D image by cooperation of multiple mobile devices. The cooperation of mobile devices brings out a special feature, cascaded view, which has a property that cutting faces of different mobile devices are relative. Besides, we also provide user another method to change location of cutting face by touch gestures. Moreover, considering user experience, we develop retrospect and annotation functions for user to return to specific cutting face state and make note on it. About implementation, we have made a prototype of this system using computed tomography image as resource of 3D image.