以監控者為中心之多攝影機平順轉場技術

Abstract

目前多攝影機監控系統通常包含大量的攝影機並且設置在大範圍的監視環境。因此，在設計監控系統方面，有一個很重要的議題是如何讓使用者了解一系列事件發生的環境。例如：當目標物橫跨多台攝影機時，我們系統不同於傳統監控系統直接將目前攝影機視角直接切換到另一台攝影機視角的方式，我們提出一個新的方法是以使用者為中心平順轉換攝影機視角，讓使用者可以很簡單的了解攝影機之間的相對關係和地理位置。在切換攝影機視角時，我們的系統會結合預先建立好的虛擬背景的三維模型以及合成的前景，然後再由虛擬攝影機成像。我們系統有一個很重要的特質就是，在攝影機所拍攝範圍只有一小部分有交集時或者是完全沒有交集時依然可以運作。這種情況之前在其他轉場系統從來沒有被處理過。此外，目前的轉場系統通常內插兩台真實攝影機的位置來決定虛擬攝影機的位置在轉場時期的位置。在我們系統，我們設計了一個決定虛擬攝影機路徑的規則，以設置虛擬攝影機的位置，取代了直接內插的方式。

Current visual surveillance systems usually include multiple cameras to monitor the activities of targets over a large area. An important issue for the guard or user using the system is to understand a series of events occurring in the environment, for example to track a target walking across multiple cameras. Opposite to the traditional systems switching the camera view from one to another directly, we propose a novel system to ease the mental effort for users to understand the geometry between real cameras and the guidance path by egocentric view transition. During the period of switching cameras, our system synthesizes the virtual views by blending the synthesized foreground texture into the pre-constructed background model and then re-projecting it to the view of virtual camera. An important property of our system is that it can be applied to the situations of where the view fields of transition cameras are not close enough or even exclusive. Such situations have never been taken into consideration in the state-of-the-art view transition techniques. In addition, current view transition systems usually linear interpolate two real cameras position to decide the virtual camera position in the period of view transition. Here, we design a rule to determine the virtual camera position instead of linear interpolation for better visual effect.