【摘要】：在用戶界面研究中,人機(jī)交互技術(shù)是當(dāng)前發(fā)展最迅速的技術(shù)之一,研究人員予以特別重視。它是一門綜合學(xué)科,與認(rèn)知學(xué)、人機(jī)工程學(xué)、心理學(xué)等學(xué)科領(lǐng)域有著密切的聯(lián)系。作為人機(jī)交互中重要的一部分,手勢(shì)識(shí)別一直以來被眾多研究者重視。特別是近幾年,隨著微軟公司的Kinect的出現(xiàn),符合人機(jī)交流習(xí)慣的手勢(shì)識(shí)別交互技術(shù)的研究變得非常活躍。按照手勢(shì)動(dòng)作分類,手勢(shì)識(shí)別研究包括兩部分:靜態(tài)手勢(shì)識(shí)別及動(dòng)態(tài)手勢(shì)識(shí)別。本課題以微軟公司提供的Kinect為手勢(shì)動(dòng)作的采集設(shè)備,對(duì)靜態(tài)手勢(shì)識(shí)別和動(dòng)態(tài)手勢(shì)識(shí)別的算法分別進(jìn)行優(yōu)化然后在虛擬場(chǎng)景中完成測(cè)試。首先,為了使手部區(qū)域分割更精確,提出一種新的手部區(qū)域分割算法。該算法通過計(jì)算軀干區(qū)域和手部區(qū)域的類間方差得到最佳分割閾值,從而提取到手部區(qū)域,再計(jì)算手部區(qū)域點(diǎn)密度最大的點(diǎn)得到掌心點(diǎn),采用相應(yīng)橢圓描述手掌區(qū)域的基礎(chǔ)上結(jié)合相應(yīng)坐標(biāo)系將手部區(qū)域細(xì)分成手掌區(qū)域、指尖區(qū)域和手臂區(qū)域。其次,針對(duì)靜態(tài)手勢(shì)識(shí)別過程中利用單特征識(shí)別時(shí)準(zhǔn)確率低的問題,提出一種基于多特征提取的手勢(shì)識(shí)別算法。此算法首先提取指尖點(diǎn)到手掌中心點(diǎn)的距離、指尖點(diǎn)到手掌平面的距離和手掌區(qū)域三種不同的手勢(shì)特征,然后應(yīng)用一個(gè)多分類的支持向量機(jī)(SVM)分類器對(duì)靜態(tài)手勢(shì)進(jìn)行分類,并在手勢(shì)數(shù)據(jù)庫中完成了算法驗(yàn)證。第三,針對(duì)動(dòng)態(tài)手勢(shì)識(shí)別過程中關(guān)節(jié)點(diǎn)獲取不準(zhǔn)確的問題,提出一種利用關(guān)節(jié)點(diǎn)可信度度量關(guān)節(jié)點(diǎn)有效性的算法。此算法通過計(jì)算關(guān)節(jié)點(diǎn)的行為可信度、運(yùn)動(dòng)學(xué)可信度和彩色圖像可信度及其可信度的特征權(quán)重,可更準(zhǔn)確獲取動(dòng)態(tài)手勢(shì)的關(guān)節(jié)點(diǎn),從而完成快速準(zhǔn)確的動(dòng)態(tài)手勢(shì)識(shí)別。最后,在基于3ds Max和Unity 3d設(shè)計(jì)的三維虛擬場(chǎng)景中完成實(shí)時(shí)檢測(cè)。結(jié)合靜態(tài)手勢(shì)和動(dòng)態(tài)手勢(shì)識(shí)別技術(shù),設(shè)計(jì)包括開始、指向、轉(zhuǎn)向、放縮、揮手及停止等手勢(shì)動(dòng)作,驅(qū)動(dòng)虛擬場(chǎng)景完成相應(yīng)功能的實(shí)時(shí)變化,驗(yàn)證了算法的有效性。
[Abstract]:In the research of user interface, human-computer interaction is one of the most rapidly developing technologies, and researchers pay special attention to it. It is a comprehensive subject and has close relation with cognitive science, ergonomics, psychology and so on. As an important part of human-computer interaction, gesture recognition has been paid attention to by many researchers. Especially in recent years, with the emergence of Microsoft Kinect, the research on gesture recognition and interaction technology, which accords with man-machine communication habits, has become very active. According to gesture classification, gesture recognition includes two parts: static gesture recognition and dynamic gesture recognition. In this paper, the Kinect provided by Microsoft is used as the acquisition device of gesture action. The algorithms of static gesture recognition and dynamic gesture recognition are optimized and tested in virtual scene. Firstly, in order to make hand region segmentation more accurate, a new hand region segmentation algorithm is proposed. The algorithm obtains the optimal segmentation threshold by calculating the variance between the torso region and the hand region, and then extracts the hand region, and then calculates the point with the highest density in the hand region to get the centerpoint. On the basis of describing the palm region with the corresponding ellipse, the hand region is subdivided into palm region, fingertip region and arm region in the corresponding coordinate system. Secondly, aiming at the problem of low accuracy when using single feature in static gesture recognition, a gesture recognition algorithm based on multi-feature extraction is proposed. The algorithm firstly extracts the distance from the fingertip to the center of the palm, the distance from the fingertip to the palm plane and three different gesture features in the palm area. Then, a multi-classification support vector machine (SVM) classifier is used to classify the static gestures. The algorithm is verified in the gesture database. Thirdly, aiming at the problem of inaccuracy of node acquisition in dynamic gesture recognition, an algorithm is proposed to measure the effectiveness of the node by using the reliability of the node. By calculating the behavioral credibility, kinematics credibility and the feature weights of the color image credibility, the algorithm can obtain the dynamic gesture nodes more accurately, so as to complete the fast and accurate dynamic gesture recognition. Finally, real-time detection is completed in a three-dimensional virtual scene based on 3ds Max and Unity 3D design. Combined with static gesture and dynamic gesture recognition technology, the design includes start, point, turn, drop, wave and stop gestures, drive the virtual scene to complete the corresponding real-time changes, and verify the effectiveness of the algorithm.
【學(xué)位授予單位】：中北大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類號(hào)】：TP391.41

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 張立志;黃菊;孫華東;趙志杰;陳麗;邢宗新;;局部特征與全局特征結(jié)合的HMM靜態(tài)手勢(shì)識(shí)別[J];計(jì)算機(jī)科學(xué);2016年S2期

2 李凱;王永雄;孫一品;;一種改進(jìn)的DTW動(dòng)態(tài)手勢(shì)識(shí)別方法[J];小型微型計(jì)算機(jī)系統(tǒng);2016年07期

3 朱娟;;手勢(shì)識(shí)別在教學(xué)中的應(yīng)用[J];信息系統(tǒng)工程;2016年06期

4 易靖國;程江華;庫錫樹;;視覺手勢(shì)識(shí)別綜述[J];計(jì)算機(jī)科學(xué);2016年S1期

5 郭曉利;楊婷婷;張雅超;;基于Kinect深度信息的動(dòng)態(tài)手勢(shì)識(shí)別[J];東北電力大學(xué)學(xué)報(bào);2016年02期

6 毛雁明;章立亮;;基于Kinect骨架追蹤技術(shù)的PPT全自動(dòng)控制方法研究[J];海南大學(xué)學(xué)報(bào)(自然科學(xué)版);2015年03期

7 談家譜;徐文勝;;基于Kinect的指尖檢測(cè)與手勢(shì)識(shí)別方法[J];計(jì)算機(jī)應(yīng)用;2015年06期

8 劉嘯宇;韓格欣;王瑞;代麗男;薄純娟;;一種基于Kinect的手勢(shì)識(shí)別系統(tǒng)[J];物聯(lián)網(wǎng)技術(shù);2015年05期

9 劉佳;鄭勇;張小瑞;Pp冬慧;陸熊;;基于Kinect的手勢(shì)跟蹤概述[J];計(jì)算機(jī)應(yīng)用研究;2015年07期

10 屈燕琴;李昕;盧夏衍;;基于表觀特征分析的手勢(shì)識(shí)別及其應(yīng)用[J];計(jì)算機(jī)工程與科學(xué);2015年01期

，

本文編號(hào)：2424269