【摘要】：三維動畫制作是計算機應用領域的重要組成部分，是隨著計算機軟件和硬件技術的發(fā)展而興起的一種技術。計算機動畫由于其真實性、精確性和可操作性，被應用到人們生活的各個領域，如娛樂、醫(yī)學、軍事和教育等。其中群體動畫受到人們越來越多的關注，成為熱點問題。群體動畫的應用非常廣泛，電影中模擬大規(guī)模戰(zhàn)爭場面、公共場所人流模擬、交通規(guī)劃、防災等。本文主要關注于動漫制作中群體動畫的路徑規(guī)劃。 由于群體行為的復雜性，群體動畫成為三維動畫制作中具有挑戰(zhàn)性的工作。傳統(tǒng)的動畫制作使用“關鍵幀”技術，不僅缺乏真實性，且當動畫角色數(shù)量很大時，角色之間或角色與環(huán)境之間交互的工作量很大。改變一個角色的動作的工作量相當于對整個動畫重新進行制作。為了解決群體動畫制作中真實性差與效率低下的問題，本文探索了一個解決方案，對群體動畫制作具有非常重要的意義。 本文對群體動畫制作的群智能方法進行了詳細的分析，對群體動畫的制作流程和實現(xiàn)方法進行了探討，設計并實現(xiàn)了一個支持自動路徑規(guī)劃的三維CA D系統(tǒng)。該系統(tǒng)結合M i c r o s o f t V i s u a l C++6.0與M a y a軟件，能夠模擬幾種常見的群體運動。 本文的主要工作是： 1）采用了一種高效的虛擬環(huán)境建模方法。增加了虛擬角色與虛擬環(huán)境的實時交互問題，同時盡可能減少虛擬環(huán)境的存儲量，提高了虛擬數(shù)據(jù)的存儲和檢索效率。采用改進的多層次環(huán)境模型。 2）設計了一種高效的虛擬角色建模方法。針對大量虛擬角色數(shù)量的、角色相似但不完全相同面臨的傳統(tǒng)手工設計方法工作量大、效率低、創(chuàng)新性不足等問題。采用基于遺傳算法的NURBS群體模型生成方法。 3）設計了一個角色進行自動路徑規(guī)劃的實現(xiàn)方法。環(huán)境中的角色進行路徑規(guī)劃以群體智能理論為主要算法，分三個步驟：選擇行為、規(guī)劃行為、實現(xiàn)行為。此方法實現(xiàn)了角色的快速自動路徑規(guī)劃，保證了動畫的效率性。 4）設計了一個基于碰撞檢測與碰撞避免的感知系統(tǒng)。提高了群體動畫的仿真度。在研究生物多種感知能力的基礎上設計了一個以觸覺感知為重點的感知系統(tǒng)。 5）在上面工作的基礎上，設計并實現(xiàn)了一個支持自動路徑規(guī)劃的三維CAD系統(tǒng) AP P C AD系統(tǒng)。本文系統(tǒng)的研究了群體動畫的實現(xiàn)方法，以生物科學、計算機科學及設計方法的交叉學科為理論基礎，以生物科學的群體智能理論為主要算法，以動漫制作為背景，設計并實現(xiàn)了一個支持自動路徑規(guī)劃的三維CAD系統(tǒng) APPCAD系統(tǒng)。APPCAD系統(tǒng)將群體動畫的制作流程分為四個步驟：環(huán)境建模、角色建模、路徑規(guī)劃、動畫制作。
[Abstract]:3D animation is an important part of computer application field. It is a kind of technology rising with the development of computer software and hardware technology. Because of its authenticity, accuracy and maneuverability, computer animation has been applied to various fields of people's life, such as entertainment, medicine, military and education. Among them, group animation attracts more and more attention and becomes a hot issue. The application of group animation is very extensive, the movie simulates the large-scale war scene, the public place people flow simulation, the transportation plan, the disaster prevention and so on. This paper mainly focuses on the path planning of group animation in animation production. Because of the complexity of group behavior, group animation has become a challenging task in 3D animation production. Traditional animation using "key frame" technology, not only lack of authenticity, but also when the number of animation characters is very large, the interaction between characters or between roles and the environment is very heavy. The amount of work required to change the action of a character is equivalent to remaking the entire animation. In order to solve the problems of poor authenticity and low efficiency in the production of group animation, this paper explores a solution, which is of great significance to the production of group animation. In this paper, the swarm intelligence method of group animation is analyzed in detail, the process and realization of group animation are discussed, and a 3D CAD system supporting automatic path planning is designed and implemented. Combined with M i c r o s o f t V i s u a l C 6.0 and M a y a software, the system can simulate several common group movements. The main work of this paper is as follows: 1) an efficient virtual environment modeling method is adopted. The problem of real-time interaction between virtual role and virtual environment is increased, the storage capacity of virtual environment is reduced as much as possible, and the efficiency of storage and retrieval of virtual data is improved. An efficient virtual role modeling method is designed by using the improved multi-level environment model. In order to solve the problems of large amount of virtual characters, similar but not identical characters, the traditional manual design methods have many problems, such as heavy workload, low efficiency and lack of innovation. The method of generating NURBS population model based on genetic algorithm is adopted. 3) A method of automatic path planning for role is designed. The main algorithm of path planning in environment is swarm intelligence theory, which is divided into three steps: selecting behavior, planning behavior, and realizing behavior. This method realizes the fast automatic path planning of characters and ensures the efficiency of animation. 4) A collision detection and collision avoidance sensing system is designed. The simulation degree of group animation is improved. Based on the research of biological multi-perception ability, a perceptual system focusing on tactile perception is designed. 5) based on the above work, A three-dimensional CAD system named AP P C AD which supports automatic path planning is designed and implemented. In this paper, the realization method of group animation is studied systematically, which is based on the interdiscipline of biology science, computer science and design method, takes the swarm intelligence theory of biology science as the main algorithm, and takes the animation production as the background. This paper designs and implements a 3D CAD system, APPCAD system, which supports automatic path planning. It divides the production process of group animation into four steps: environment modeling, role modeling, path planning and animation making.
【學位授予單位】：電子科技大學
【學位級別】：碩士
【學位授予年份】：2012
【分類號】：TP311.52

【參考文獻】

相關期刊論文 前10條

1 謝菠蓀;;虛擬聽覺在虛擬現(xiàn)實、通信及信息系統(tǒng)的應用[J];電聲技術;2008年01期

2 謝菠蓀;;虛擬聽覺環(huán)境的原理、進展和問題[J];電聲技術;2008年11期

3 宋永超;張勇軍;蔡澤祥;楊銀國;朱鷹屏;;一種電力系統(tǒng)無功優(yōu)化的新型混合優(yōu)化算法[J];電力系統(tǒng)保護與控制;2011年24期

4 劉豐,莊越挺,羅忠祥,潘云鶴;基于多自主智能體的群體動畫創(chuàng)作[J];計算機研究與發(fā)展;2004年01期

5 張雯栶;高守平;;改進共享策略的簡單群搜索優(yōu)化算法[J];計算機工程與科學;2011年07期

6 史豐;陳福民;王憶源;;一種基于人工生命的群體動畫[J];計算機應用;2005年S1期

7 董剛;趙龍;田尊華;;基于智能體的群體動畫創(chuàng)作方法的研究與實現(xiàn)[J];計算機仿真;2005年12期

8 陳泓娟,班曉娟,涂序彥,盧漢清;計算機動畫的人工生命方法研究——人工魚的自繁衍模型[J];自動化學報;2003年06期

9 李子蓉;;美、日、韓動漫產(chǎn)業(yè)發(fā)展經(jīng)驗及對我國的啟示[J];世界地理研究;2006年04期

10 胡瑞波;徐人平;向極鼎;;虛擬觸覺設計Freeform在產(chǎn)品開發(fā)設計中的應用研究[J];陜西科技大學學報(自然科學版);2011年05期

，

本文編號：2171965