應用於人體姿態辨識之自我學習乏析規則分類系統

智慧型監控系統在未來的科學研究領域當中，將會是一個值得令人討論的技術。其應用層面相當之廣泛，不管在居家看護、公共場所安全檢測…等，經由這類的技術發展，不僅可以降低成本還可以進而改善人為疏失的問題。在本篇論文當中，將會著重在居家看護系統分析與建立的部分。而在居家看護當中我門所關切的是人類動作行為異常的判斷，也就是說，隨著老年化的社會成長，越來越多的老年人會待在家中，而當這些老人跌倒甚至發生危險時，我們將會透過這樣的系統提醒正在遠方工作的兒女，讓他們可以及時的做出解救的動作。而在本論文中，除了異常姿態的辨識，我們也將辨識一般日常行為中的人體姿態。我們首先利用CCD攝影機擷取環境影像，並且利用背景去除的影像處理方式，將移動中的物體擷取出來，並對此物體作簡單的特徵擷取。藉由得取的這些參數，利用本篇提及的適應性自我學習乏析規則分類系統，產生乏析規則並且利用這些規則判斷人體姿態。本系統建立了一個常態的乏析規則庫，並將此應用在人體姿態的分類，最後達成智慧型辨識的能力。

關鍵字

居家看護系統；人體姿態辨識；乏析規則；背景去除

並列摘要

Intelligent video surveillance system is discussed for years and applied in many areas, like home care systems, security in the public place and so on. Due to this technique development, it can lower down the product cost and decrease the error judgment by human. In this thesis, we will focus on the home care systems, it means that we hope to establish a system can analysis the human activities automatically. As the growth of the elderly, more and more people can not take care of their parents or the elderly all the time. It manes that we want to build a system have the ability to watch the elderly and alarm the people when they have some abnormal behavior, like falling down. In our concept we first want to recognize the human postures, so we set a CCD camera to grab the image contained the elderly and environment. And then we separate the people from the environment using the background subtraction method. After the object is subtracted, some parameters from the silhouette will be extracted and viewed as the input of our classified system. As for the classified system, we use the adaptive fuzzy rule-based classified system. It can generate the fuzzy rules automatically according to the feature parameters we extract, and give a good performance in classifying the postures.

並列關鍵字

human activities ； background subtraction ； adaptive fuzzy rule based classified system

參考文獻

[1]. F. Buccolieri, C. Distance and A. Leone, “Human Posture Recognition Using Active Contours and Radial Basis Function Neural Network”, IEEE conference on Advanced Video and Signal Based Surveillance, 2005. AVSS 2005, 15-16 Sept. 2005 Page(s):213–218 Digital Object Identifier 10.1109/AVSS.

[2]. Ronald Poppe and Mannes Poel, “Comparison of Silhouette Shape Descriptors for Example-based Human Pose Recovery”, Proceedings of the 7th International Conference on Automatic Face and Gesture Recognition (FGR’06), 2006.

[3]. J.K. Aggarwal and Q. Cai, ”Human motion analysis: a review”, Nonrigid and Articulated Motion Workshop, 1997. Proceedings, IEEE, Page(s):90 – 102,16 June 1997.

[4]. G. Johansson, “Visual motion perception”, Sci. Am. 232(6), 1975, 76-88.

[5]. Hironobu Fujiyoshi, Alan J. Lipton and Takeo Kanade, “Real-Time Human Motion Analysis by Image Skeletonization”, IEICE TRANS. INF. & SYST., Vol. E87-D, No.1 January 2004.

國際替代計量

應用於人體姿態辨識之自我學習乏析規則分類系統

全文下載

主題瀏覽