Full metadata record
DC FieldValueLanguage
dc.contributor.author陳政輝en_US
dc.contributor.authorCheng-Hui Chenen_US
dc.contributor.author莊仁輝en_US
dc.contributor.authorJen-Hui Chuangen_US
dc.date.accessioned2014-12-12T02:23:00Z-
dc.date.available2014-12-12T02:23:00Z-
dc.date.issued1999en_US
dc.identifier.urihttp://140.113.39.130/cdrfb3/record/nctu/#NT880394073en_US
dc.identifier.urihttp://hdl.handle.net/11536/65573-
dc.description.abstract廣義圓柱體 (GC) 架構是物體形狀描述方法中重要的方法之一,本論文提出一個新的方法來求得一個廣義圓柱體。首先,我們利用一個基於廣義位能場模型來搜尋物體骨架的演算法 [3],透過種子點的選擇,求出物體的GC軸。從GC軸上均勻取出取樣點之後,再利用一個計算物體截面的演算法以求得描述這個物體的輪廓線。使用廣義位能場的原因,是因為它可以有效率地計算出GC軸,並且不需要就特定的物體形狀類別作考慮,較具一般性。從實驗結果可以看出,我們所提出的演算法可以有效的產生描述多種形狀物體的廣義圓柱體。zh_TW
dc.description.abstractGeneralized cylinder (GC) scheme is one of the important shape description methods in computer vision. This paper proposes a new approach to derive a generalized cylinder description of a 3D object. The approach is based on the skeletonization algorithm proposed in [3] using a generalized potential field model, and finds the GC axis of an object by selecting alternative seed points. Cross sections at sampling points uniformly distributed along the GC axis are then obtained by a simple algorithm. Such a potential-based approach is quite general in the sense that it is applicable to a variety of shape classes. The experimental results demonstrate that our approach can find GC axes effectively for objects of various shapes.en_US
dc.language.isozh_TWen_US
dc.subject廣義圓柱體zh_TW
dc.subject3D物體形狀描述zh_TW
dc.subjectgeneralized cylinderen_US
dc.subject3D shape descriptionsen_US
dc.title基於位能場的廣義圓柱體研究zh_TW
dc.titleThe Research on Potential-Based Generalized Cylinderen_US
dc.typeThesisen_US
dc.contributor.department資訊科學與工程研究所zh_TW
Appears in Collections:Thesis