本文選題：手勢(shì)分組 + 隱性馬爾科夫��； 參考：《哈爾濱工業(yè)大學(xué)》2017年碩士論文

【摘要】：使用可穿戴式設(shè)備進(jìn)行手勢(shì)識(shí)別正在快速地成為一個(gè)研究熱點(diǎn),被廣泛地應(yīng)用在了行為檢測(cè)、手語(yǔ)識(shí)別以及人機(jī)交互方面。今天,隨著微機(jī)電系統(tǒng)(Micro Electromechanical System,MEMS)的發(fā)展使得生產(chǎn)更小、更輕便的傳感器和設(shè)備成為了可能,這些設(shè)備可以穿戴在人身上用來(lái)檢測(cè)人的行為甚至一些更小的肢體動(dòng)作。當(dāng)用戶(hù)在資源受限的穿戴式設(shè)備上執(zhí)行實(shí)際的手勢(shì)識(shí)別的時(shí)候,需要考慮識(shí)別精度和算法的時(shí)間復(fù)雜度。目前許多的手勢(shì)識(shí)別算法已經(jīng)被提出并采用了。隱性馬爾科夫模型(Hidden Markov Model,HMM)是目前用的比較多的手勢(shì)識(shí)別算法,HMM最早多被用于語(yǔ)音識(shí)別,由于手勢(shì)序列和語(yǔ)音序列的相似性,因此被廣泛用于手勢(shì)識(shí)別當(dāng)中,且可以達(dá)到比較高的識(shí)別精度。由于HMM的高運(yùn)算復(fù)雜度,導(dǎo)致當(dāng)用戶(hù)在資源受限的移動(dòng)設(shè)備上使用HMM來(lái)進(jìn)行手勢(shì)識(shí)別的時(shí)候,達(dá)不到實(shí)時(shí)反應(yīng)的效果,用戶(hù)體驗(yàn)差,需要進(jìn)行改進(jìn)。HMM的運(yùn)算復(fù)雜度直接與需要識(shí)別的數(shù)據(jù)集規(guī)模、觀(guān)察序列的長(zhǎng)度以及狀態(tài)數(shù)目成正比,減少這三個(gè)參數(shù)的值能夠降低運(yùn)算復(fù)雜度,但是識(shí)別精度也相應(yīng)的變低了。因此有必要找出一種方法:在維持識(shí)別精度的前提下盡量減低算法運(yùn)算復(fù)雜度,使其適應(yīng)移動(dòng)終端的運(yùn)算能力。針對(duì)上述提出的問(wèn)題,本文提出了一種通過(guò)對(duì)手勢(shì)進(jìn)行分組來(lái)降低識(shí)別算法的運(yùn)算復(fù)雜度以及通過(guò)為每個(gè)組設(shè)定不同的HMM來(lái)保持甚至提高識(shí)別精度的方法。此方法包含三部分內(nèi)容:手勢(shì)分組,組模型的建立,以及每個(gè)組里面手勢(shì)模型的建立。手勢(shì)分組使用基于K-means++的方法;組模型使用基于表格的方法;手勢(shì)模型使用HMM,同一個(gè)組內(nèi)的HMM具有相似的結(jié)構(gòu),不同組具有不同的結(jié)構(gòu)。為了驗(yàn)證分組手勢(shì)識(shí)別方法的有效性,本文定義了12種手勢(shì),這些手勢(shì)考慮到了不同的形狀、方向以及重復(fù)性,具有代表性。然后,本研究搭建了一個(gè)數(shù)據(jù)采集平臺(tái),這個(gè)平臺(tái)包括可穿戴式的硬件平臺(tái)以及電腦端上位機(jī)平臺(tái)。通過(guò)這個(gè)平臺(tái),本研究采集了大量的手勢(shì)數(shù)據(jù)來(lái)驗(yàn)證分組手勢(shì)識(shí)別方法的有效性。實(shí)驗(yàn)結(jié)果表明:在不損失識(shí)別精度的前提下,和標(biāo)準(zhǔn)的HMM相比,本文提出的對(duì)手勢(shì)進(jìn)行分組手勢(shì)識(shí)別的方法的運(yùn)算復(fù)雜度大大降低,驗(yàn)證了方法的可行性。
[Abstract]:Gesture recognition using wearable devices is becoming a research hotspot and has been widely used in behavior detection, sign language recognition and human-computer interaction. Today, with the development of MEMS Micro Electromechanical Systems, it is possible to produce smaller, lighter sensors and devices that can be worn on people to detect human behavior and even smaller body movements. When the user performs actual gesture recognition on a wearable device with limited resources, the recognition accuracy and the time complexity of the algorithm should be considered. At present, many gesture recognition algorithms have been proposed and adopted. Hidden Markov Model HMMM (Hidden Markov Model) is one of the most widely used gesture recognition algorithms at present. Due to the similarity between gesture sequences and speech sequences, hmm is widely used in gesture recognition. And it can achieve higher recognition accuracy. Due to the high computational complexity of HMM, when users use HMM for gesture recognition on resource-constrained mobile devices, the real-time response is not achieved and the user experience is poor. The computational complexity of the improved. Hmm is directly proportional to the size of the dataset to be identified, the length and the number of states of the observation sequence. Reducing the values of these three parameters can reduce the computational complexity, but the recognition accuracy is also reduced accordingly. Therefore, it is necessary to find a way to reduce the computational complexity of the algorithm under the premise of maintaining the recognition accuracy, so that it can adapt to the computing ability of the mobile terminal. To solve the above problems, this paper proposes a method to reduce the computational complexity of the recognition algorithm by grouping the gestures and to maintain or even improve the recognition accuracy by setting different HMM for each group. The method consists of three parts: gesture grouping, group modeling, and gesture modeling within each group. Gesture grouping uses K-means based approach, group model uses tabular approach, gesture model uses HMMs, and HMM in the same group has similar structure and different groups have different structures. To verify the effectiveness of the grouping gesture recognition method 12 gestures are defined which take into account different shapes directions and repeatability and are representative. Then, a data acquisition platform is built, which includes wearable hardware platform and PC platform. Through this platform, this study collected a lot of gesture data to verify the effectiveness of the grouping gesture recognition method. The experimental results show that the computational complexity of the proposed method is greatly reduced compared with the standard HMM without loss of recognition accuracy, and the feasibility of the method is verified.
【學(xué)位授予單位】：哈爾濱工業(yè)大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類(lèi)號(hào)】：TP391.41

【參考文獻(xiàn)】

相關(guān)期刊論文 前9條

1 郝旭歡;常博;郝旭麗;;MEMS傳感器的發(fā)展現(xiàn)狀及應(yīng)用綜述[J];無(wú)線(xiàn)互聯(lián)科技;2016年03期

2 王月明;趙士偉;張如彩;;基于人機(jī)交互的虹膜圖像采集系統(tǒng)設(shè)計(jì)[J];中國(guó)安防;2014年17期

3 王杰鋒;周治平;苗敏敏;;移動(dòng)終端手勢(shì)識(shí)別中DTW匹配算法研究[J];計(jì)算機(jī)工程與應(yīng)用;2015年13期

4 陳意;楊平;陳旭光;;一種基于加速度特征提取的手勢(shì)識(shí)別方法[J];傳感技術(shù)學(xué)報(bào);2012年08期

5 荊雷;馬文君;常丹華;;基于動(dòng)態(tài)時(shí)間規(guī)整的手勢(shì)加速度信號(hào)識(shí)別[J];傳感技術(shù)學(xué)報(bào);2012年01期

6 王萬(wàn)良;楊經(jīng)緯;蔣一波;;基于運(yùn)動(dòng)傳感器的手勢(shì)識(shí)別[J];傳感技術(shù)學(xué)報(bào);2011年12期

7 洪淑月;施曉鐘;徐皓;;改進(jìn)的小波變換HMM語(yǔ)音識(shí)別算法[J];浙江師范大學(xué)學(xué)報(bào)(自然科學(xué)版);2011年04期

8 張雪鳳;張桂珍;劉鵬;;基于聚類(lèi)準(zhǔn)則函數(shù)的改進(jìn)K-means算法[J];計(jì)算機(jī)工程與應(yīng)用;2011年11期

9 任程;戴樹(shù)嶺;;基于數(shù)據(jù)手套的逼真虛擬手的實(shí)現(xiàn)[J];系統(tǒng)仿真學(xué)報(bào);2008年22期

相關(guān)博士學(xué)位論文 前1條

1 申慧敏;基于頭部電磁信息反演的人機(jī)交互系統(tǒng)控制指令提取方法研究[D];浙江大學(xué);2015年

相關(guān)碩士學(xué)位論文 前2條

1 陳文;基于加速度傳感器的智能終端手勢(shì)識(shí)別關(guān)鍵技術(shù)研究[D];國(guó)防科學(xué)技術(shù)大學(xué);2011年

2 楊筆鋒;基于改進(jìn)訓(xùn)練算法的HMM語(yǔ)音識(shí)別技術(shù)研究[D];湖南大學(xué);2010年

，

本文編號(hào)：1877723