查詢結果分析
相關文獻
- 以動態擷取系統之光標數為聚類之類別數為基礎的人體動作混合方法
- 模糊資料之聚類分析
- 應用灰色預測與灰色聚類於航空公司航線運量預測與形態設計
- 模糊與灰色評估方法之比較--以高速鐵路技術型式之評估為例
- 應用模糊聚類分析理論模擬具分散式處理理念而移植於行動臺之新交遞控制演算法則
- 以模糊C平均值聚類法則為基礎之M準位相位位移鍵調變型態分類研究
- Classification of Water Masses and Sound-Scattering Layer Biomass in the Waters off Northeastern Taiwan Using a Fuzzy Clustering Method
- A Fuzzy Document Retrieval System based on Concept Networks and Cluster Analysis
- 臺灣中部地區空氣品質預測之探討
- 臺灣地區銀行業顧客滿意度之研究--自組織映射類神經網路之應用
頁籤選單縮合
題名 | 以動態擷取系統之光標數為聚類之類別數為基礎的人體動作混合方法=Human Motion Blending by Using the Markers of Motion Capture as Clusters |
---|---|
作者 | 王榮英; 杜宗翰; 徐方正; | 書刊名 | 資訊科學應用期刊 |
卷期 | 6:1 2010.06[民99.06] |
頁次 | 頁105-118 |
分類號 | 312.49 |
關鍵詞 | 動作擷取; 動作混合; 聚類; 線性內插; Motion capture; Motion blending; Clustering; Linear interpolation; |
語文 | 中文(Chinese) |
中文摘要 | 本研究提出一個能快速及平順的混合各種動作之動作混合系統,使用者可以根據動畫的需求,直接將兩種或兩種以上的動作做混合,以達到所需要的效果。主要的貢獻在我們接以動態擷取設備之光標點數目做為聚類演算法之類別數目,明確的解決了以往以聚類方法進行動作混合之研究時,需任意給定一門檻值而產生類別數之缺點。另一方面,為簡化與加快動作混合的效率,我們利用聚類演算法之特性,分別尋找出欲混合之兩個動作中,據起承轉合之代表性動作之關鍵影格,以其進行線性內插獲得兩動作間之混合。至於時間軸扭曲(time warpping)方面,我們則以4段的內插法來取代。為驗證本研究所提出之動作混合方法,我們以一系列的動作混合實作驗證之,結果顯示我們所提的方法可快速及平順的完成兩 相異動間動作間之動作混合。 |
英文摘要 | In this paper we present a general method for motion blending that could be used to smoothly blend two motion clips and could automatically select the transition key frames. Animators can use this approach to produce a new move sequence. The main contribution of our study is that we directly utilize the amount of makers in motion capture equipment as the clusters in clustering method to avoid the drawback of assigning the thresholds of fuzzy clusters by hand. In order to simplify and accelerate the efficiency of motion blending, each motion clip is selected only one frame, as the key frame, to represent all time frames of that motion. This key frame retains the important characteristic of that motion clip. We use the characteristics of k-mean clustering algorithm to find out the key frame of each motion clip. After that, we do the motion blending by linear interpolation between the two representational key frames. We also use the 4 stages of linear interpolation to replace the time wrapping process. To evaluate the effectiveness of the proposed method, we implemented a series of action mixed experiments. The results show that our method is effective in smoothly blending between two motion sequences. |
本系統之摘要資訊系依該期刊論文摘要之資訊為主。