English  |  正體中文  |  简体中文  |  Items with full text/Total items : 43312/67235
Visitors : 2090501      Online Users : 14
RC Version 5.0 © Powered By DSPACE, MIT. Enhanced by NTU/NCHU Library IR team.

Please use this identifier to cite or link to this item: http://nchuir.lib.nchu.edu.tw/handle/309270000/111708

標題: 3D立體視覺於手勢切割之應用
Hand Gesture Segmentation Using 3D Stereo Vision
作者: 鄭喬勻
Cheng, Chiao-Yun
Contributors: 廖俊睿;郭伯臣
陶金旭
中興大學
關鍵字: stereo vision;depth information;labeling algorithm;skin detection;morphological algorithm;wrist detection
立體視覺;深度資訊;標籤演算法;Parzen Window;膚色偵測;形態學演算法;手腕偵測;最小平方法線
日期: 2013
Issue Date: 2012-09-04 16:18:42 (UTC+8)
Publisher: 通訊工程研究所
摘要: 本篇論文主要是研究手勢切割。我們可將這個研究分為手部區域的切割與手腕偵測兩個部份。首先是完整手勢影像的取得,透過雙鏡頭立體視覺攝影機獲得彩色影像、灰階影像與深度資訊後,將影像與深度資訊利用標籤演算法、Parzen Window的膚色偵測演算法以及形態學演算法進行處理。經過這些處理我們感興趣的完整手勢便可以從複雜背景中分割出來。再來就是如何準確且有效率地找到手腕的部份,在手腕偵測的部份,我們計算手部最大內接圓來決定手臂的上、下界以避免不同的手勢對於最小平方線在運算上造成的影響,並且求得手臂的最小平方法線與邊緣輪廓。利用找出手勢影像邊緣輪廓合適直線的方式,計算出手臂輪廓與合適直線的誤差,找到最大誤差點即為手腕的位置。根據垂直最小平方線的斜率與最大誤差點來精確的推算出手腕分割線並且切割出完整的手勢。最後經過與其他切割方式比較之後,可以發現我們提出的手勢分割方法所求得的手腕分割割線更加接近使用者實際的手腕位置,並且在主觀視覺上擁有較好的切割效果。
In this thesis, we study the hand gesture segmentation with a complex background. We can roughly divide the research into two parts: the hand/arm segmentation and the wrist detection. The first one is to obtain the complete hand/arm image region. We can acquire color images, gray-level images, and depth information through the stereo camera. Then we make use of the labeling, skin detection, and morphological algorithms to segment the hand/arm region from the complex background. The second part is to accurately locate the wrist. We calculate the maximum inner circle of the hand/arm region to decide the upper and lower thresholds for the wrist and to avoid the adverse influence due to different hand gestures. Then we obtain the edge contour of the arm. The contour is used to fit piecewise line segments. The point with the maximum fitting deviation is considered as the wrist position. Finally, the hand region can be segmented by removing the region below the wrist line. According to the experimental results, our approach can obtain visually better segmentation and the segmentation line which is nearer the position of real human wrist compared with other segmentation methods.
Appears in Collections:[依資料類型分類] 碩博士論文

Files in This Item:

File SizeFormat
index.html0KbHTML456View/Open


 


學術資源

著作權聲明

本網站為收錄中興大學學術著作及學術產出,已積極向著作權人取得全文授權,並盡力防止侵害著作權人之權益。如仍發現本網站之數位內容有侵害著作權人權益情事者,請權利人通知本網站維護人員,將盡速為您處理。

本網站之數位內容為國立中興大學所收錄之機構典藏,無償提供學術研究與公眾教育等公益性使用。

聯絡網站維護人員:wyhuang@nchu.edu.tw,04-22840290 # 412。

DSpace Software Copyright © 2002-2004  MIT &  Hewlett-Packard  /   Enhanced by   NTU/NCHU Library IR team Copyright ©   - Feedback