基于圖優(yōu)化的單目視覺SLAM技術(shù)研究
本文選題:機(jī)器人同步定位與建圖 + 單目視覺; 參考:《華中科技大學(xué)》2016年碩士論文
【摘要】:機(jī)器人同步定位與建圖(Simultaneous Localization and Mapping,簡稱SLAM)技術(shù)是機(jī)器人實(shí)現(xiàn)定位的主流研究方法。隨著傳感器和計(jì)算機(jī)視覺等技術(shù)的發(fā)展,視覺傳感器越來越多的被裝備到機(jī)器人上。單個(gè)攝像頭具有性價(jià)比高,結(jié)構(gòu)簡單,適用范圍廣等優(yōu)點(diǎn),因此單目視覺SLAM方法受到越來越多的關(guān)注和研究。單目視覺SLAM方法主要分為基于濾波器的方法和基于圖優(yōu)化的方法,前者計(jì)算復(fù)雜度較高,無法滿足實(shí)時(shí)性要求,因此,本文主要研究基于圖優(yōu)化的單目視覺SLAM方法。并行跟蹤,局部建圖和回環(huán)檢測的單目視覺SLAM方法(ORB-SLAM)是目前比較完備和可靠的方法,本文對其進(jìn)行了研究和實(shí)驗(yàn),并針對其中的一些問題進(jìn)行了改進(jìn)。ORB特征檢測算法只對一定范圍內(nèi)的旋轉(zhuǎn)和尺度變化具有不變性,因此本文設(shè)計(jì)了評估實(shí)驗(yàn)對各種特征檢測算法進(jìn)行評測,找到了性能最佳的適合在單目視覺SLAM中使用的特征檢測算法。特征的數(shù)目和分布對跟蹤效果和相機(jī)姿態(tài)估計(jì)會(huì)產(chǎn)生影響,穩(wěn)定的特征數(shù)目和均勻的特征分布會(huì)使算法更加魯棒,本文設(shè)計(jì)了一種可以控制特征點(diǎn)提取的數(shù)目,并且使特征點(diǎn)在圖像中盡量均勻分布的算法。在地圖的初始化和局部建圖中生成新的地圖點(diǎn)時(shí),ORB-SLAM使用的是線性三角化方法(Least Square Method,簡稱LS算法),該方法生成的地圖點(diǎn)的位置有可能不準(zhǔn)確,本文將其改進(jìn)為迭代的三角化方法(Iterative Least Square Method,簡稱It-LS算法)。因?yàn)楸疚母倪M(jìn)的特征檢測算法使用的特征描述子為FREAK,因此將改進(jìn)后的SLAM方法簡稱為FREAK-SLAM。針對以上改進(jìn),本文設(shè)計(jì)了相關(guān)的實(shí)驗(yàn)來進(jìn)行驗(yàn)證。本文在室內(nèi)、室外環(huán)境下對FREAK-SLAM進(jìn)行了實(shí)驗(yàn),實(shí)驗(yàn)表明FREAK-SLAM跟蹤到的特征點(diǎn)的數(shù)目更加穩(wěn)定,特征點(diǎn)的分布更加均勻,且相機(jī)的定位精度較ORB-SLAM有所提高,室內(nèi)環(huán)境下平均提高了0.18cm,室外環(huán)境下平均提高了1.4m,提高的百分比約為10%。
[Abstract]:Simultaneous localization and mapping (slam) technology is the main research method for robot localization. With the development of sensors and computer vision, more and more vision sensors are equipped on robots. Single camera has many advantages, such as high cost performance, simple structure and wide range of application, so the single vision slam method has been paid more and more attention and research. Monocular vision slam method is mainly divided into filter based method and graph based optimization method. The former has high computational complexity and can not meet the real-time requirements. Therefore, this paper mainly studies Monocular vision slam method based on graph optimization. Monocular vision slam (Orb-slam) is a complete and reliable method for parallel tracking, local mapping and loop detection. The improved .Orb feature detection algorithm is only invariant to the rotation and scale change in a certain range, so this paper designs an evaluation experiment to evaluate the various feature detection algorithms. The best feature detection algorithm for monocular slam is found. The number and distribution of features will affect the tracking effect and camera attitude estimation, and the stable number of features and uniform feature distribution will make the algorithm more robust. In this paper, we design a method to control the number of feature points extracted. And the algorithm to make the feature points distribute evenly in the image as far as possible. When generating new map points in map initialization and local mapping, ORB-SLAM uses a linear triangulation method called least Square method (LS algorithm for short). The location of map points generated by this method may not be accurate. In this paper, the iterative triangulation method is improved as iterative least Square method (It-LS algorithm for short). Because the feature descriptor used in the improved feature detection algorithm is FREAK, the improved slam method is referred to as FREAK-SLAM. In view of the above improvements, this paper designed relevant experiments to verify. The experiment of FREAK-SLAM in indoor and outdoor environment shows that the number of feature points tracked by FREAK-SLAM is more stable, the distribution of feature points is more uniform, and the positioning accuracy of camera is higher than that of ORB-SLAM. The average increase was 0.18 cm in indoor environment and 1.4 m in outdoor environment. The percentage of increase was about 10 cm.
【學(xué)位授予單位】:華中科技大學(xué)
【學(xué)位級別】:碩士
【學(xué)位授予年份】:2016
【分類號】:TP391.41;TP242
【參考文獻(xiàn)】
相關(guān)期刊論文 前4條
1 王忠立;趙杰;蔡鶴皋;;大規(guī)模環(huán)境下基于圖優(yōu)化SLAM的后端優(yōu)化方法[J];哈爾濱工業(yè)大學(xué)學(xué)報(bào);2015年07期
2 顧照鵬;劉宏;;單目視覺同步定位與地圖創(chuàng)建方法綜述[J];智能系統(tǒng)學(xué)報(bào);2015年04期
3 張國良;姚二亮;岳亞南;孫一杰;;一種魯棒閉環(huán)的增量式Graph SLAM算法[J];電光與控制;2015年07期
4 王忠立;趙杰;蔡鶴皋;;大規(guī)模環(huán)境下基于圖優(yōu)化SLAM的圖構(gòu)建方法[J];哈爾濱工業(yè)大學(xué)學(xué)報(bào);2015年01期
相關(guān)博士學(xué)位論文 前2條
1 梁明杰;機(jī)器人認(rèn)知地圖創(chuàng)建關(guān)鍵技術(shù)研究[D];華南理工大學(xué);2014年
2 孟旭炯;基于單目視覺的同時(shí)定位與地圖重建算法研究[D];浙江大學(xué);2010年
相關(guān)碩士學(xué)位論文 前6條
1 胡衡;基于單目視覺的SLAM方法研究[D];新疆大學(xué);2015年
2 張彥珍;基于g2o的SLAM后端優(yōu)化算法研究[D];西安電子科技大學(xué);2014年
3 張珊;基于PCA-SURF的單目視覺SLAM算法[D];中國海洋大學(xué);2014年
4 翟佳龍;基于KAZE特征的單目視覺SLAM研究[D];西安電子科技大學(xué);2014年
5 李捐;基于單目視覺的移動(dòng)機(jī)器人SLAM問題的研究[D];哈爾濱工業(yè)大學(xué);2013年
6 江龍;基于SURF特征的單目視覺SLAM技術(shù)研究與實(shí)現(xiàn)[D];南京理工大學(xué);2012年
,本文編號:2012149
本文鏈接:http://sikaile.net/kejilunwen/ruanjiangongchenglunwen/2012149.html