一個影片的內容可以視為一連串的影格，而每一個影格可以經由轉換以一個符號影像(symbolic image)來表示。在先前許多相關的研究中，根據符號影像中物件間空間關係之影片相似尋取是一個影片尋取的重要方法。然而，此方法並未明確的考慮到物件在影片中相鄰影格間的空間關係變化及移動資訊。 在本論文中，我們定義在相鄰影格間，因各物件不同的移動向量所發生空間關係的變化稱為空間事件(Spatial Event)。當多個空間事件同時發生時，以字串的方式表示，稱為空間事件字串(Spatial Event String, 簡稱SES)。我們提出計算兩SES間之相似度方法。將查詢的SES序列與參考影片之SES序列做比較，並根據SES間之相似度，我們提出在參考影片之SES序列中找出與查詢影片相似的片段的方法。因此，即可進行以空間事件為基礎的影片相似尋取。 A video can be viewed as a sequence of frames. Each frame can be transformed into a symbolic image. In the previous research, similarity retrieval of video based on the spatial relationship between objects in the symbolic image is an important method. The change of the spatial relationship between objects and the information about the movement of the objects in the video are not considered directly and precisely. In this paper, the spatial event is defined as the change of the spatial relationship between objects due to the movements of the objects from a frame to its next. When multiple events occur, the spatial events are represented as a string, named the spatial event string (SES). We propose a method to calculate the similarity between two SESs. Similarity retrieval of video can be performed by comparing the query SES sequence with the SES sequence of the reference video. We proposed a method to locate the segment in the reference video similar to the query SES sequence based on the similarity between SESs. Hence, Similarity retrieval of video based on spatial events can be performed.
