發(fā)布時(shí)間：2018-03-30 23:36

  本文選題：Kinect　切入點(diǎn)：姿勢識(shí)別　出處：《沈陽工業(yè)大學(xué)》2017年碩士論文

【摘要】：在機(jī)器視覺和圖像處理領(lǐng)域,關(guān)于人體姿勢的識(shí)別已經(jīng)成為一項(xiàng)重要的課題,并且在人機(jī)交互、虛擬現(xiàn)實(shí)和智能視頻監(jiān)控等領(lǐng)域得到廣泛的應(yīng)用。然而目前仍有諸多問題沒有得到很好的解決,影響了計(jì)算機(jī)對(duì)于人體行為的理解。由于普通攝像機(jī)只能獲取到二維圖像,但二維信息到三維信息的重建會(huì)丟失很多重要數(shù)據(jù),影響動(dòng)作識(shí)別的精度。盡管科研人員設(shè)計(jì)了多種圖像重構(gòu)算法,但是仍無法避免光照、紋理遮擋等影響。而Kinect傳感器使用一種新的獲取圖像的方式,它通過一對(duì)紅外攝像頭捕獲到帶有空間距離的深度圖像,并且在深度圖像的基礎(chǔ)上提取出含有三維坐標(biāo)信息的骨骼數(shù)據(jù)流。但是Kinect沒有給出姿勢識(shí)別的高級(jí)函數(shù),原因在于人體動(dòng)作千變?nèi)f化,很難構(gòu)建出一套通用的模型進(jìn)行識(shí)別。為了提升動(dòng)作識(shí)別的效果,本文使用Kinect傳感器來獲取到人體20個(gè)骨骼關(guān)節(jié)點(diǎn)的三維坐標(biāo),并且根據(jù)人體姿態(tài)的特征,以關(guān)節(jié)點(diǎn)的相對(duì)距離和角度序列為特征參數(shù)。在靜態(tài)姿勢的識(shí)別中,通過特征向量對(duì)樣本集進(jìn)行訓(xùn)練,并使用KNN算法作為分類器對(duì)姿勢進(jìn)行識(shí)別。在動(dòng)作評(píng)價(jià)中,在分析了運(yùn)動(dòng)特征序列的時(shí)間特性以后,采用線性回歸的方法對(duì)樣本曲線進(jìn)行訓(xùn)練,使用最小二乘法擬合出一條最佳角度曲線作為標(biāo)準(zhǔn)模板,在考慮到曲線之間時(shí)間序列長短不一的問題,通過DTW算法對(duì)不同長度的關(guān)節(jié)角度曲線進(jìn)行匹配,并且通過定義一套公式對(duì)動(dòng)作進(jìn)行評(píng)價(jià),以曲線之間DTW差值作為實(shí)驗(yàn)參數(shù),最終將評(píng)價(jià)方法應(yīng)用到動(dòng)作打分的體感游戲中。本文通過實(shí)驗(yàn)分析了相對(duì)距離和關(guān)節(jié)點(diǎn)角度作為動(dòng)作識(shí)別特征向量的可行性。選取簡氏太極拳其中的4式作為靜態(tài)姿勢識(shí)別對(duì)象,實(shí)驗(yàn)結(jié)果證明通過該方法進(jìn)行姿態(tài)識(shí)別可以獲得較高的識(shí)別率。然后又對(duì)動(dòng)作評(píng)價(jià)的實(shí)驗(yàn)數(shù)據(jù)進(jìn)行分析,在總結(jié)8個(gè)角度的DTW差值樣本點(diǎn)分布規(guī)律之后,定義一套公式對(duì)動(dòng)作進(jìn)行評(píng)價(jià),并設(shè)計(jì)動(dòng)作評(píng)估系統(tǒng)論證該動(dòng)作評(píng)價(jià)方法的合理性。由于評(píng)價(jià)公式中的基數(shù)和因子會(huì)隨著動(dòng)作的變換而不斷重新計(jì)算,增加了該評(píng)價(jià)方法的復(fù)雜度,接下來的工作是完善動(dòng)作評(píng)價(jià)公式的各項(xiàng)參數(shù)使評(píng)價(jià)方法具有更優(yōu)的效率和適應(yīng)性。
[Abstract]:In the field of machine vision and image processing, recognition of human posture has become an important issue, and in human-computer interaction, Virtual reality and intelligent video surveillance are widely used. However, there are still many problems that have not been solved well, which affect the understanding of human behavior by computer. But the reconstruction of two-dimensional to three-dimensional information can lose a lot of important data and affect the accuracy of motion recognition. Although researchers have designed a variety of image reconstruction algorithms, but still can not avoid lighting, The Kinect sensor uses a new way to capture images, which capture depth images with spatial distances through a pair of infrared cameras. On the basis of the depth image, the skeletal data stream with three-dimensional coordinate information is extracted. But Kinect does not give a high-level function of posture recognition, because the human body's actions vary greatly. It is very difficult to construct a universal model for recognition. In order to improve the effect of motion recognition, we use Kinect sensor to get the three-dimensional coordinates of 20 skeletal joints, and according to the characteristics of human posture, The relative distance and angle sequence of the node are taken as the characteristic parameters. In the recognition of static pose, the sample set is trained by the feature vector, and the posture is recognized by using KNN algorithm as the classifier. After analyzing the time characteristics of motion feature series, the linear regression method is used to train the sample curve, and the least square method is used to fit an optimal angle curve as the standard template. Considering the difference of time series between curves, the joint angle curve of different length is matched by DTW algorithm, and the action is evaluated by defining a set of formulas. The difference of DTW between curves is taken as the experimental parameter. Finally, the evaluation method is applied to the body feeling game of action scoring. The feasibility of using relative distance and the angle of gate node as the feature vectors of action recognition is analyzed experimentally in this paper. Four of the four forms of Taijiquan are selected as static. State posture recognition object, The experimental results show that the attitude recognition rate can be obtained by this method. Then, the experimental data of motion evaluation are analyzed, and the distribution of DTW difference sample points from 8 angles is summarized. A set of formulas is defined to evaluate the action, and a motion evaluation system is designed to demonstrate the rationality of the evaluation method. The complexity of the evaluation method is increased. The next work is to improve the parameters of the action evaluation formula to make the evaluation method more efficient and adaptive.
【學(xué)位授予單位】：沈陽工業(yè)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

【相似文獻(xiàn)】

相關(guān)期刊論文 前10條

1 彭衍琳;;金融行業(yè)客戶滿意度的模糊綜合評(píng)價(jià)方法[J];中國金融電腦;2007年05期

2 袁運(yùn)祥;三峽工程對(duì)生態(tài)與環(huán)境影響的綜合評(píng)價(jià)方法[J];計(jì)算機(jī)應(yīng)用;1988年02期

3 黃西川;高新技術(shù)項(xiàng)目開發(fā)優(yōu)化度評(píng)價(jià)方法研究[J];科研管理;1994年03期

4 杜銳,隋洋;企業(yè)技術(shù)含量的概念及評(píng)價(jià)方法研究[J];科學(xué)管理研究;1997年02期

5 柏雪飛;;計(jì)算機(jī)基礎(chǔ)課程的教學(xué)評(píng)價(jià)方法探討[J];學(xué)苑教育;2013年02期

6 修永輝;;國家產(chǎn)業(yè)競爭力的信息評(píng)價(jià)方法[J];情報(bào)雜志;2006年01期

7 張其翔;趙欣艷;呂廷杰;;電信企業(yè)的商業(yè)模式及其評(píng)價(jià)方法[J];電信科學(xué);2007年01期

8 王磊;施榮華;;企業(yè)信息化多層次模糊綜合評(píng)價(jià)方法[J];系統(tǒng)工程;2012年02期

9 唐納德·E·里格斯;楊柳;;圖書館戰(zhàn)略規(guī)劃的評(píng)價(jià)方法[J];圖書館;1986年06期

10 張家，

本文編號(hào)：1688190