本文選題：癱瘓康復 + 視頻信息　； 參考：《濟南大學》2017年碩士論文

【摘要】：在目前社會中,經濟發(fā)展使得人們生活水平逐漸提高,與此同時,因為車禍、中風等導致重度癱瘓的人員也逐漸增多。這不僅給患者的生理以及心理造成巨大損害,其治療與康復工作也給家庭帶來了較為沉重的負擔。重度癱瘓患者中有一類比較特殊的情況:喪失語言能力且四肢失去活動能力,但大腦意識清醒,仍然具有頭部活動能力與眼嘴張閉能力。據(jù)山東中醫(yī)藥大學附屬醫(yī)院粗略統(tǒng)計,這類患者占到了重度癱瘓患者總人數(shù)的10%左右。如果有辦法能獲取這類患者的真實需求,使患者能夠自主控制康復機械臂或康復床等康復機械達到照料自己的目的,降低監(jiān)護成本和負擔,減少患者痛苦,同時也會解決很多家庭問題。這類患者具有清醒的自主意識和正常的頭部以及面部動作,因此可以采用攝像頭采集患者頭部以及眼嘴動作視頻信息,利用情感計算相關技術對這些動作進行識別與翻譯,得到患者真實需求(喝水、坐起、翻身等)的方法。本課題具體研究內容是基于情感計算的重度癱瘓患者人機需求交互。其中“情感計算”指的是對于頭部動作(左右搖頭、上下擺頭)與面部動作(眼睛睜閉、嘴巴張閉)的識別。其中“人機需求交互”方式是:對于患者的常用、緊急的需求(喝水、求救等)由特定動作組合的方式直接發(fā)出;對于普通的需求(坐起、翻身等),患者通過動作組合控制系統(tǒng)光標移動,模擬鼠標操作選擇需求按鈕向康復床發(fā)出動作需求。主要工作包括:(1)創(chuàng)建了重度癱瘓患者層次需求模型,根據(jù)該模型設計了通過情感控制模擬鼠標操作的交互方式與交互界面;(2)將情感計算中的頭部動作與面部動作識別算法進行了研究與改進,通過區(qū)域化的頭部動作識別算法、基于灰度曲面的眼睛睜閉動作識別算法以及基于亮度的嘴部動作識別算法獲得了較好的動作識別效果;(3)開發(fā)了一套基于上述模型與算法的重度癱瘓患者人機需求交互系統(tǒng),并取得了較好的測試效果。本文的創(chuàng)新點在于:(1)首次采用情感計算的方式進行重度癱瘓患者需求獲取;(2)構建了重度癱瘓患者層次需求模型,進而開發(fā)了相應的人機交互系統(tǒng)。本研究成果結合該項目其他參與單位所研制的康復床等機械設備,使患者可以自主完成生活起居以及部分康復動作,大幅度減少看護人員工作量,提高患者生活與康復質量。
[Abstract]:In the present society, the economic development makes people's living standard improve gradually, at the same time, because of the traffic accident, the stroke and so on causes the serious paralysis person also to increase gradually. This not only causes great damage to the patient's physiology and psychology, but also brings heavy burden to the family. Among the patients with severe paralysis, there are some special cases: loss of speech ability and limb inactivity, but the brain is conscious and still has the ability of head movement and mouth opening and closing. According to the statistics of affiliated Hospital of Shandong University of traditional Chinese Medicine, these patients account for about 10% of the total number of patients with severe paralysis. If there is a way to obtain the real needs of such patients so that they can control the rehabilitation arm or the rehabilitation machinery such as the rehabilitation bed to achieve the purpose of taking care of themselves, reducing the cost and burden of guardianship, and reducing the suffering of the patients, It also solves a lot of family problems. These patients have a clear sense of autonomy and normal head and facial movements, so they can use a camera to capture video information of head and mouth movements, and use emotional computing techniques to recognize and translate these movements. Get the patient's real needs (drink, sit up, turn over, etc.). The specific content of this research is based on emotional calculation of severe paralysis patients with human-computer interaction. "emotional computing" refers to the recognition of head movements (left and right head shaking, up-and-down head swinging) and facial movements (eyes open, mouth open). The "man-machine demand interaction" approach is: for patients commonly used, urgent needs (water, help, etc.) are sent out directly by a specific combination of actions; for ordinary needs (sit up, etc.) Turning over and so on, the patient controls the cursor movement of the system through the action combination, and simulates the mouse operation selection demand button to issue the action demand to the rehabilitation bed. The main work includes the creation of a hierarchical demand model for severely paralyzed patients. According to this model, the interactive mode and interface of simulating mouse operation by emotion control are designed. The recognition algorithm of head action and facial movement in emotion calculation is studied and improved, and the recognition algorithm of head action is regionalized. The recognition algorithm based on gray surface and the recognition algorithm of mouth movement based on luminance have obtained better action recognition effect. A set of man-machine demand interaction system of severe paralysis patients based on the above model and algorithm has been developed. Good test results have been obtained. The innovation of this paper is that: (1) for the first time, emotional calculation is used to obtain the needs of patients with severe paralysis. (2) the hierarchical demand model of patients with severe paralysis is constructed, and the corresponding human-computer interaction system is developed. The results of this study combined with the rehabilitation bed and other mechanical equipment developed by other participating units of the project enable the patients to complete their daily living and part of the rehabilitation activities on their own, greatly reduce the workload of nursing staff, and improve the quality of patients' life and rehabilitation.
【學位授予單位】：濟南大學
【學位級別】：碩士
【學位授予年份】：2017
【分類號】：TP11

【參考文獻】

相關期刊論文 前10條

1 黃建;李文書;高玉娟;;人臉表情識別研究進展[J];計算機科學;2016年S2期

2 姚鵬飛;盛鑫軍;郭偉超;華磊;朱向陽;;基于表面肌電信號與近紅外光譜技術聯(lián)合解碼的仿人假手控制系統(tǒng)[J];中國康復醫(yī)學雜志;2016年05期

3 任安虎;劉貝;;基于Adaboost的人臉識別眨眼檢測[J];計算機與數(shù)字工程;2016年03期

4 曲斯偉;宋為群;;腦-機接口技術在卒中患者康復中的研究進展[J];中國腦血管病雜志;2016年03期

5 張毅;廖巧珍;羅元;;融合二階HOG與CS-LBP的頭部姿態(tài)估計[J];智能系統(tǒng)學報;2015年05期

6 方又方;喻洪流;官龍;易金花;張穎;石萍;;基于肌電觸發(fā)的上肢康復訓練機器人的實現(xiàn)[J];上海理工大學學報;2015年04期

7 胡進;侯增廣;陳翼雄;張峰;王衛(wèi)群;;下肢康復機器人及其交互控制方法[J];自動化學報;2014年11期

8 唐云祁;孫哲南;譚鐵牛;;頭部姿勢估計研究綜述[J];模式識別與人工智能;2014年03期

9 楊啟志;曹電鋒;趙金海;;上肢康復機器人研究現(xiàn)狀的分析[J];機器人;2013年05期

10 譚民;王碩;;機器人技術研究進展[J];自動化學報;2013年07期

相關會議論文 前1條

1 董力賡;陶霖密;徐光yP;;頭部姿態(tài)和動作的識別與理解[A];第三屆和諧人機環(huán)境聯(lián)合學術會議（HHME2007）論文集[C];2007年

相關博士學位論文 前1條

1 史小華;坐/臥式下肢康復機器人研究[D];燕山大學;2014年

相關碩士學位論文 前7條

1 趙琳;基于人臉識別的疲勞駕駛監(jiān)控方法研究[D];長春工業(yè)大學;2015年

2 楊偉健;基于視覺和肌電信息融合的智能輪椅人機接口技術研究及應用[D];杭州電子科技大學;2015年

3 李蕾;增強現(xiàn)實應用中的頭部動作識別技術研究[D];北方工業(yè)大學;2014年

4 劉軍;復雜環(huán)境下駕駛員眼睛定位及眼睛狀態(tài)識別算法研究[D];華南理工大學;2014年

5 劉小燮;腦機交互結合功能性電刺激康復訓練新技術對慢性期腦卒中大腦可塑性的影響[D];中國人民解放軍醫(yī)學院;2014年

6 梁海燕;基于Kinect動作驅動的三維細微面部表情實時模擬[D];燕山大學;2013年

7 楊景旭;利用Kinect估計人體頭部姿態(tài)[D];南京理工大學;2012年

，

本文編號：1956799