基于無人機(jī)平臺的目標(biāo)檢測與人機(jī)交互算法研究
發(fā)布時(shí)間：2018-08-28 11:37
【摘要】：近年來,隨著無人機(jī)控制技術(shù)的日趨成熟和人工智能技術(shù)的蓬勃發(fā)展,多旋翼無人機(jī)除了執(zhí)行傳統(tǒng)的航拍、植保、戰(zhàn)地偵查等任務(wù)之外,還被賦予了主動(dòng)跟蹤目標(biāo),自主目標(biāo)識別等智能功能。目前,多旋翼無人機(jī)大部分的應(yīng)用還是集中在民用娛樂方面,而無人機(jī)的人機(jī)交互方式仍主要依賴于較為復(fù)雜的遙控器控制,遙控器較大的體積妨礙了小型多旋翼無人機(jī)的便攜性,遙控器控制的操縱不便性阻礙了無人機(jī)使用人群的推廣。因此,需要設(shè)計(jì)一種不依賴于外接設(shè)備的智能人機(jī)交互方法。人體姿態(tài)信息較為明顯且對人較為親和,通過計(jì)算機(jī)視覺與人工智能技術(shù),對人體姿態(tài)信息進(jìn)行識別,就可以完成人機(jī)交互任務(wù)。首先,針對無人機(jī)在空中飛行時(shí)機(jī)械振動(dòng)造成拍攝畫面抖動(dòng)的情況,采用中心區(qū)域模板匹配的方法,在算法上實(shí)現(xiàn)了對無人機(jī)機(jī)械振動(dòng)的補(bǔ)償,使得畫面較為穩(wěn)定并易于后續(xù)算法處理。接著,對圖像中的人體目標(biāo)進(jìn)行檢測�？紤]使用了兩種解決方案,分別用于地面站和機(jī)上處理情況。第一種是使用結(jié)構(gòu)復(fù)雜,運(yùn)算量大的卷積神經(jīng)網(wǎng)絡(luò)檢測器Faster R-CNN完成候選區(qū)域的生成,目標(biāo)種類的分類與人體目標(biāo)位置的定位,該方法準(zhǔn)確度高,能夠適應(yīng)復(fù)雜的背景環(huán)境,但運(yùn)算量龐大,無法在機(jī)載嵌入式計(jì)算平臺上實(shí)現(xiàn)實(shí)時(shí)性。另一種是使用前景信息提取的方法生成候選框,再使用結(jié)構(gòu)較為簡單的卷積神經(jīng)網(wǎng)絡(luò)完成分類任務(wù),該方法需要先對場景信息進(jìn)行建模,但運(yùn)算量小,可以在機(jī)載嵌入式平臺上實(shí)現(xiàn)實(shí)時(shí)性。然后,對檢測到的人的姿態(tài)信息進(jìn)行識別。為了確定檢測到的人是否為需要進(jìn)行人機(jī)交互的對象,設(shè)計(jì)了一個(gè)揮手動(dòng)作檢測器作為人體姿態(tài)信息檢測器的開關(guān),增強(qiáng)了系統(tǒng)的安全性。對通過了揮手動(dòng)作檢測的對象,設(shè)計(jì)了4種明顯的姿勢,無人機(jī)通過對姿勢種類的識別完成人機(jī)交互任務(wù)。為了提高運(yùn)行速度并充分利用已處理信息,將框選到的目標(biāo)前景信息直接作為特征輸入多層全連接神經(jīng)網(wǎng)絡(luò)進(jìn)行識別。最后,將算法移植到了機(jī)載嵌入式設(shè)備上并進(jìn)行了飛行實(shí)驗(yàn)驗(yàn)證。結(jié)果表明,系統(tǒng)能夠在保持較高識別準(zhǔn)確率的同時(shí)實(shí)現(xiàn)了機(jī)上處理的實(shí)時(shí)性,能夠很好的完成人機(jī)交互任務(wù)。
[Abstract]:In recent years, with the increasing maturity of UAV control technology and the vigorous development of artificial intelligence technology, multi-rotors UAVs have been given active tracking targets in addition to performing traditional aerial photography, plant protection, field reconnaissance and other tasks. Autonomous target recognition and other intelligent functions. At present, most of the applications of multi-rotor UAV are concentrated in the field of civil entertainment, and the man-machine interaction mode of UAV still mainly depends on the more complex remote control. The large volume of remote control hinders the portability of small multi-rotor UAV, and the inconvenience of remote control hinders the popularizing of UAV users. Therefore, it is necessary to design an intelligent human-computer interaction method which is independent of external devices. The attitude information of human body is obvious and affable to human being. Through computer vision and artificial intelligence technology, the human posture information can be recognized, and the human-computer interaction task can be accomplished. First of all, aiming at the situation that the mechanical vibration of UAV flying in the air causes the shooting picture jitter, the center region template matching method is used to compensate the UAV mechanical vibration in the algorithm. Make the picture more stable and easy to follow up algorithm processing. Then, the human body target in the image is detected. Two solutions are considered, one for earth station and the other for machine processing. The first is the use of convolutional neural network detector (Faster R-CNN), which has complex structure and large computation, to generate candidate regions, classify target types and locate human body targets. This method has high accuracy and can adapt to complex background environment. However, it is difficult to implement real-time on the airborne embedded computing platform. The other is to use the method of extracting foreground information to generate candidate boxes, and then to use convolution neural network with simple structure to complete the classification task. This method needs to model the scene information first, but the computation is small. Real-time performance can be realized on the airborne embedded platform. Then, the attitude information of the detected person is recognized. In order to determine whether the detected person is the object of human-computer interaction, a wave action detector is designed as the switch of the human attitude information detector, which enhances the security of the system. For the objects that have been detected by waving, four kinds of postures are designed, and the UAV realizes the human-computer interaction task by recognizing the types of gestures. In order to improve the running speed and make full use of the processed information, the target foreground information selected by the frame is directly input as a feature into the multi-layer fully connected neural network for recognition. Finally, the algorithm is transplanted to the airborne embedded equipment and flight experiment is carried out. The results show that the system can achieve real-time processing while maintaining high recognition accuracy, and can accomplish human-computer interaction task well.
【學(xué)位授予單位】：哈爾濱工業(yè)大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2017
【分類號】：TP11;V279
本文編號：2209283
[Abstract]:In recent years, with the increasing maturity of UAV control technology and the vigorous development of artificial intelligence technology, multi-rotors UAVs have been given active tracking targets in addition to performing traditional aerial photography, plant protection, field reconnaissance and other tasks. Autonomous target recognition and other intelligent functions. At present, most of the applications of multi-rotor UAV are concentrated in the field of civil entertainment, and the man-machine interaction mode of UAV still mainly depends on the more complex remote control. The large volume of remote control hinders the portability of small multi-rotor UAV, and the inconvenience of remote control hinders the popularizing of UAV users. Therefore, it is necessary to design an intelligent human-computer interaction method which is independent of external devices. The attitude information of human body is obvious and affable to human being. Through computer vision and artificial intelligence technology, the human posture information can be recognized, and the human-computer interaction task can be accomplished. First of all, aiming at the situation that the mechanical vibration of UAV flying in the air causes the shooting picture jitter, the center region template matching method is used to compensate the UAV mechanical vibration in the algorithm. Make the picture more stable and easy to follow up algorithm processing. Then, the human body target in the image is detected. Two solutions are considered, one for earth station and the other for machine processing. The first is the use of convolutional neural network detector (Faster R-CNN), which has complex structure and large computation, to generate candidate regions, classify target types and locate human body targets. This method has high accuracy and can adapt to complex background environment. However, it is difficult to implement real-time on the airborne embedded computing platform. The other is to use the method of extracting foreground information to generate candidate boxes, and then to use convolution neural network with simple structure to complete the classification task. This method needs to model the scene information first, but the computation is small. Real-time performance can be realized on the airborne embedded platform. Then, the attitude information of the detected person is recognized. In order to determine whether the detected person is the object of human-computer interaction, a wave action detector is designed as the switch of the human attitude information detector, which enhances the security of the system. For the objects that have been detected by waving, four kinds of postures are designed, and the UAV realizes the human-computer interaction task by recognizing the types of gestures. In order to improve the running speed and make full use of the processed information, the target foreground information selected by the frame is directly input as a feature into the multi-layer fully connected neural network for recognition. Finally, the algorithm is transplanted to the airborne embedded equipment and flight experiment is carried out. The results show that the system can achieve real-time processing while maintaining high recognition accuracy, and can accomplish human-computer interaction task well.
【學(xué)位授予單位】：哈爾濱工業(yè)大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2017
【分類號】：TP11;V279
【參考文獻(xiàn)】
相關(guān)期刊論文 前4條
1 李漢森;;無人機(jī)在行業(yè)應(yīng)用領(lǐng)域的最新發(fā)展[J];中國安防;2017年05期
2 陳晶;許軍;吳微露;毛毅;;全球鷹無人機(jī)的升級及試驗(yàn)情況[J];飛航導(dǎo)彈;2015年05期
3 王暉娟;蔣琪;;2012珠海航展拾萃[J];飛航導(dǎo)彈;2012年11期
4 劉鑫;許華榮;胡占義;;基于GPU和Kinect的快速物體重建[J];自動(dòng)化學(xué)報(bào);2012年08期
相關(guān)博士學(xué)位論文 前1條
1 丁衛(wèi);基于超小型無人機(jī)的地面目標(biāo)實(shí)時(shí)圖像跟蹤[D];上海大學(xué);2007年
相關(guān)碩士學(xué)位論文 前5條
1 高翔;基于機(jī)載嵌入式平臺的無人機(jī)視覺輔助自主降落[D];哈爾濱工業(yè)大學(xué);2016年
2 王冉;基于深度卷積神經(jīng)網(wǎng)絡(luò)的人體姿勢估計(jì)研究[D];電子科技大學(xué);2016年
3 吳杰;基于深度學(xué)習(xí)的手勢識別研究[D];電子科技大學(xué);2015年
4 張志飛;小型無人直升機(jī)視覺定位與跟蹤系統(tǒng)的設(shè)計(jì)與研究[D];浙江大學(xué);2013年
5 鐘佳朋;四旋翼無人機(jī)的導(dǎo)航與控制[D];哈爾濱工業(yè)大學(xué);2010年
，本文編號：2209283
本文鏈接：http://sikaile.net/kejilunwen/zidonghuakongzhilunwen/2209283.html
最近更新
教材專著
