本論文，使用了原先幫助機器人做同時定位與地圖創建(Simultaneous Localization and Mapping ,SLAM)分析的裝置，它的名稱是UTM-30LX，因為它可以做即時的地圖建置，所以幫助人在周遭環境在看不清楚情形下，偵測周圍的活動範圍，再根據人和裝置間的相對位置，讓我們辨識出可以行走的方向和距離，本論文提出一個類似火場的情境，把生還者用一個箱子取代，放在角落，使得路徑長會變短，根據電腦接收裝置偵測到的數據做分析，看有沒有箱子擋住。
    首先我們先根據兩個場景做例子，判斷要擺放幾個位置可以建置整個圖形，之後每擺放一次，就取一次區塊的數值，先把這些數值由極座標的方式配置，再轉換一般的2維座標數值繪出，根據實際場景擺放位置間相對的座標長度，用拼出的方式建置出一個完整的場景圖。
    取出完整的場景圖後，就可以得知各個擺放位置間長度和座標點，在後面兩章根據這樣的條件，先讓兩個場景各別建置一個的火場情境，每個場景角落在擺一個箱子當生還者，
前一章節我們用「資料庫路徑長位置比對法」的方式，先建置完所有門口到角落路徑長，之後根據擺放箱子後路徑長的變化值和位置，插入原先建置好的資料庫位置對應路徑值取代舊有的路徑值，之後跟原先的資料庫比較就得知哪個位置有箱子擋住，之後繪出新的路徑長，和計算出到箱子的路徑方位和長度，藉由無線網路的方式把資訊由電腦傳送到手機給測試人員做實際場景測試。
   再於花費的時間過長，本論文提出一個新的演算法，先寫一個在270度大小下偵測三個方位障礙物的演算法，之後根據固定距離的擺置方式，及時測到障礙物後，立即繪出到障礙物的路近長和方位，同時也把資訊透過無線網路的方式由電腦傳給手機給測試人員做實際場景測試，經過實際的錄影方式呈現後，測試人員皆可順利找到箱子，由於前一章和後一章所花費的時間差異大，整個建置的時間前一章大約要3小時，後一章只要20分鐘，相較之下後一章在火場的建置和救援應用下價值較高。

關鍵詞 : 同時定位與地圖創建

In this thesis, the use of the device was originally to help robots to do the analysis of SLAM (Simultaneous the Localization and Mapping), it calls  the UTM-30LX, because it can be done to reconstruct the map immediately, so it can help people not to see clearly the situation in the surrounding environment , around the detection range of activities, according to these two relative positions configuring between people and device, let us identify the direction and distance of walking and therefore propose a similar fire situations, it replaces survivors by a group box,and put on the corner  to make the path length becomes shorter,final the computer accepts the data which the device detects for analysis to see whether there is a box to block or not.
   At first  We based on two scenarios for example,determine to how many positions place can reconstruct entire map , and then placed at a time, take the value of one block, configure these values by the way of the polar coordinates, and then convert the general 2-dimensional coordinate values plotted according to the actual scene placed in position between the length of the relative coordinate and original point, so we use a method to reconstruct a complete scene graph by puzzle.
   Obtain the complete scene graph, you can see that the length and coordinates of points between  placement of each block.Next two chapters under such conditions, let the two scenes individually build a fire situation, corner of each scene put a box instead of survivors.
   Previous section, we use such law which compares path length and position of database,first reconstruct completely all from the entrance to corner of the path length , and then is based on placing in a box to produce the value of path length and location changes, insert the pre-reconstruct location of database correspond to the value of path length to replace the old one, compared with the original database to know where the location of the box to block, and then draw the new path length, and calculate the orientation and length of the path of the box, the information is transfer from the computer to the phone by wireless network to make testers do the actual scenario testing. 
Second chapter based on the previous chapter, it takes the shortcomings of the build time is too long to do improve, the first to write an algorithm to detect the obstacle in the three azimuth within 270 degree size, followed by the placement of a fixed distance, After detecting obstacle immediately , and drawn the path length and the position between the door and the obstacle, the information is also transferred from the computer to the phone by wireless network to make tester do the actual scenario testing , testers can be successfully found the box after the actual recording presented, due to spend time very different between the previous chapter and after the chapter,the entire build time of the previous chapter need to 3 hours ,and after the one need to 20 minutes, Based on the results of comparison, after the one applied to rescue and reconstruct in fire situation which is higher worth.

目錄 VI
圖目錄VII
表目錄IX
第一章 緒論	1
1.1 研究動機與目的	1
1.2 回顧SLAM相關背景知識	2
1.3 論文章節內容簡介	3
第二章	實際重建電腦化技術簡介	4
2.1 UTM -30LX 技術規格	4
2.2 UTM 30LX 地圖建置原理	7
第三章	電腦化技術地圖重建實測原理	10
3.1 技術應用構想	10
3.1.1 技術應用方法	10
3.1.2 技術應用流程	13
3.2 實際場景分析測試	15
3.2.1 場景一地圖建置	15
3.2.2 場景一分析測試	21
3.2.3 場景二地圖建置	25
3.2.4 場景二分析測試	36
第四章	快速電腦化地圖重建實測原理	41
4.1 改善技術應用構想	41
4.1.1 改善技術應用方法	42
4.1.2 改善技術應用流程	44
4.2 實際場景分析測試	45
4.2.1 場景一實際情境	45
4.2.3 場景二實際情境	49
4.2.4 場景二分析測試	51
第五章	結論與未來展望	55
5.1 比較兩種方式運作流程完後所需時間	55
5.2 結論和未來情境	56
參考文獻	62
圖目錄
圖2.1  Hokuyo UTM-30LX	4
圖2.2  UTM-30LX 掃描取樣的範圍	5
圖2.3  UTM-30LX 控制信號的掃描方式	7
圖2.4  UTM-30LX 控制信號的時序圖	7
圖2.5  UTM-30LX 資料取樣的流程圖	8
圖3.1  應用情境構想	10
圖3.2  技術流程圖	13
圖3.3  場景一擺放配置圖	15
圖3.4  UML-30LX在位置1掃描到的範圍	16
圖3.5  UML-30LX在位置1拍攝實體圖	16
圖3.6  UML-30LX在位置2掃描到的範圍	17
圖3.7  UML-30LX在位置2拍攝實體圖	17
圖3.8  UML-30LX在位置3掃描到的範圍	18
圖3.9  UML-30LX在位置3拍攝實體圖	18
圖3.10  UML-30LX在位置4掃描到的範圍	19
圖3.11  UML-30LX在位置4拍攝實體圖	19
圖3.12  場景一所有位置拼完圖後的建置圖	20
圖3.13  場景一測試的情境	21
圖3.14  場景一座標點和路徑的規劃	21
圖3.15  場景一模擬好的所有可行路徑	22
圖3.16  場景一模擬偵測到障礙物路徑	22
圖3.17  場景一實體圖障礙物擺置情境	23
圖3.18  場景一電腦計算方位和步伐的結果	23
圖3.19  場景一手機端接收訊息接收結果	24
圖3.20  場景二UML-30LX擺放配置圖	25
圖3.21  UML-30LX在位置1掃描到的範圍	26
圖3.22  UML-30LX在位置1拍攝實體圖	26
圖3.23  UML-30LX在位置2掃描到的範圍	27
圖3.24  UML-30LX在位置2拍攝實體圖	27
圖3.25  UML-30LX在位置3掃描到的範圍	28
圖3.26  UML-30LX在位置3拍攝實體圖	29
圖3.27  UML-30LX在位置4掃描到的範圍	29
圖3.28  UML-30LX在位置4拍攝實體圖	30
圖3.29  UML-30LX在位置5掃描到的範圍	30
圖3.30  UML-30LX在位置5拍攝實體圖	31
圖3.32  UML-30LX在位置6拍攝實體圖	32
圖3.34  UML-30LX在位置7拍攝實體圖	33
圖3.35  UML-30LX在位置8掃描到的範圍	34
圖3.36  UML-30LX在位置8拍攝實體圖	34
圖3.37  場景二所有位置拼圖完後的建置圖	35
圖3.38  場景二測試情境	36
圖3.39  場景二座標點和路徑的規劃	37
圖3.40  場景二模擬好的所有可行路徑	37
圖3.41  場景二模擬偵測到障礙物路徑	38
圖3.42  場景二實體圖障礙物擺置情境	38
圖3.43  場景二電腦計算方位和步伐的結果	39
圖3.44  場景二手機端接收訊息接收結果	39
圖4.1  障礙物判斷方式	42
圖4.3  場景一情境圖	45
圖4.4  場景一實體測試擺置拍攝圖	46
圖4.5  場景一實體擺置偵測障礙物拍攝圖	46
圖4.6  場景一到達障礙物路徑軌跡模擬圖	47
圖4.7  場景一電腦端分析障礙物位置和行走步伐資訊	47
圖4.8  場景一手機端接收資訊	48
圖4.9  場景二情境圖	49
圖4.10  場景二實體拍攝圖	50
圖4.11  場景二實體擺置偵測障礙物拍攝圖	50
圖4.12  場景二到達障礙物路徑軌跡模擬圖	51
圖4.13  場景二電腦端分析障礙物位置和行走步伐資訊	52
圖4.14  場景二手機端接收資訊	52
圖5.1  兩種情境建置方法各自所需花費的平均時間	55
圖5.2  未來的應用於雲端運算做建置分析的情境	58
圖5.3 平躺者拍攝情境圖	59
圖5.4 平躺者偵測模擬圖	59
圖5.5 側躺者拍攝情境圖	60
圖5.6 側躺者偵測模擬圖	61
表目錄
表2.1 Hokuyo UTM-30LX Specifications	5
表3.1場景一實境錄影測試網址	25
表3.2場景二實境錄影測試網址	40
表3.3場景一和場景二建置方法表(以場景一擺放方位為0度)	40
表4.1場景一實境錄影測試網址	49
表4.2場景二實境錄影測試網址	54

[1].	Liang Zhang, Xujiong Meng, Yaowu Chen, ' Unscented                  Transform for SLAM Using Gaussian Mixture Model with Particle Filter,' 2009 International Conference on Electronic Computer Technology, pp.12-17,2009
[2].	L.D.L. Perera, A. Melkumyan, E. Nettleton, ' On the Linear and Nonlinear Observability Analysis of the SLAM Problem, ' Proceedings of the 2009 IEEE International Conference on Mechatronics,April 2009
[3].	Haiwei Dong, Zhiwei Luo, Weidong Chen, ' Sparsing of Information Matrix for Practical Application of a Robot's SLAM, ' 2009 IEEE International Conference on Robotics and Automation, pp.4031-4036, May 12-17,2009
[4].	Jonathan Klippenstein, Hong Zhang, ' Performance Evaluation of Visual SLAM Using Several Feature Extractors, " The 2009 IEEE/RSJ International Conference on Intelligent Robots and Systems,pp.1574-1581,October 11-15
[5].	Gert Kootstra, Lambert R. B. Schomaker, ' Using Symmetrical Regions of Interest to Improve Visual SLAM, ' The 2009 IEEE/RSJ International Conference on Intelligent Robots and Systems, pp.930-935 , October 11-15, 2009
[6].	Wolfram Burgard , Cyrill Stachniss, Giorgio Grisetti, Bastian Steder, Rainer K‥ummerle, Christian Dornhege, Michael Ruhnke, Alexander Kleiner, Juan D. Tard’os, ' A Comparison of SLAM Algorithms Based on a Graph of Relations, ' A Comparison of SLAM Algorithms Based on a Graph of Relations, ' The 2009 IEEE/RSJ International Conference on Intelligent Robots and Systems,pp.2089-2095 October 11-15, 2009
[7].	St’ephane Magnenat, Valentin Longchamp, Michael Bonani, Michael Bonani, Michael Bonani, Paolo Germano, Hannes Bleuler, Francesco Mondada, " Affordable SLAM through the Co-Design of Hardware and Methodology," 2010 IEEE International Conference on Robotics and Automation,pp.5395-5401, May 3-8, 2010
[8].	Brian Williams, Ian Reid, " On Combining Visual SLAM and Visual Odometry, ' 2010 IEEE International Conference on Robotics and Automation, pp.3494-3500, May 3-8, 2010

[9].	Luca Carlone, Miguel Kaouk Ng, Jingjing Du, Basilio Bona, Marina Indri, " Rao-Blackwellized Particle Filters Multi Robot SLAM with Unknown Initial Correspondences and Limited Communication," 2010 IEEE International Conference on Robotics and Automation,pp.243-249, May 3-8, 2010.
[10].	V. Petridis, N. Zikos, " L-SLAM: reduced dimensionality FastSLAM algorithms, ' 2010 IEEE
[11].	Jeong-Gwan Kang, Su-Y ong An, Se-Young Oh, " Modified Neural Network Aided EKF based SLAM for Improving an accuracy of the Feature Map, " 2010 IEEE
[12].	Shoudong Huang, Yingwu Lai, Udo Frese, Gamini Dissanayake, " How far is SLAM from a linear least squares problem?, " The 2010 IEEE/RSJ International Conference on Intelligent Robots and Systems, pp.3011-3016, October 18-22, 2010
[13].	L.D.L. Perera, E. Nettleton, " On Stochastically Observable Directions of the Estimation Theoretic, " The 2010 IEEE/RSJ International Conference on Intelligent Robots and Systems,pp.4324-4331, October 18-22, 2010.
[14].	Jung-Suk Oh, Kwee-Bo Sim, " International Conference on Control, Automation and Systems 2010, " International Conference on Control, Automation and Systems 2010, pp. 1993-1996 , Oct. 27-30, 2010.
[15].	Kenta Kitajima, Hiroaki Masuzawa, Jun Miura, Junji Satake, " A Hierarchical SLAM for Uncertain Range Data, " 2010 IEEE International Conference on Multisensor Fusion and Integration for Intelligent Systems, pp.144-149, Sept. 5-7, 2010.
[16].	Egor Reentov, Geoffrey G. Messier , Sebastian Magierowski," Evaluating Wireless Network Effects for SLAM Robot Map Making, " publication in the IEEE Globecom 2010 proceedings, 2010 IEEE.
[17].	Jeong-Gwan Kang, Won-Seok Choi, Su-Yong An, Se-Young Oh, " Augmented EKF based SLAM method for Improving the Accuracy of the Feature Map, " The 2010 IEEE/RSJ International Conference on Intelligent Robots and Systems,pp.3725-3731 October 18-22, 2010.


[18].	Seo-Yeon Hwang, Jae-Bok Song, " Monocular Vision-Based SLAM in Indoor Environment Using Corner, Lamp, and Door Features From Upward-Looking Camera, ' IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 58,pp.4804-4812, OCTOBER 2011.
[19].	Bastien Vincke,  Abdelhafid Elouardi, Alain Lambert, "Design and Evaluation of an Embedded System Based SLAM Applications," SI International 2010, pp.224-229, 2010 IEEE.
[20].	John Mullane, Ba-Ngu Vo, Martin D. Adams, Ba-Tuong Vo,"A Random-Finite-Set Approach to Bayesian SLAM, " IEEE TRANSACTIONS ON ROBOTICS, VOL. 27, NO. 2,pp. 268-282,
APRIL 2011.
[21].	Adizul Ahmad, Shoudong Huang, Jianguo Jack Wang, Gamini Dissanayake, " A New State Vector for Range-Only SLAM, " 2011 Chinese Control and Decision Conference,pp3404-3409,2011 IEEE.
[22].	Bo Zhao, Xiaorui Zhu," Fast Large-scale SLAM with Improved Accuracy in Mobile Robot ", Proceedings of the 2010 IEEE International Conferenceon Robotics and Biomimetics ,pp.791-796, 
December 14-18, 2010.
[23].	Gurkan Tuna, Kayhan Gulez, v. Cagri Gungor," Communication Related Design Considerations of WSN-aided Multi-Robot SLAM ", Proceedings of the 2011 IEEE Intemational Conference on Mechatronics ,pp.493-498, April 13-15, 2011.
[24].	Henning Lategahn, Andreas Geiger, Bernd Kitt, " Visual SLAM for Autonomous Ground Vehicles," 2011 IEEE International Conference on Robotics and Automation Shanghai International Conference Center, pp. 1732-1737 , May 9-13, 2011
[25].	Chen-Han Hsiao, Chieh-Chih Wang, " Achieving Undelayed Initialization in Monocular SLAM with Generalized Objects Using Velocity Estimate-based Classification," 2011 IEEE International Conference on Robotics and Automation Shanghai International Conference Center ,pp.4060-4066. May 9-13, 2011.
[26].	Nikos Zikos, Vassilios Petridis, " L-SLAM: reduced dimensionality FastSLAM with unknown data association, " 2011 IEEE International Conference on Robotics and Automation Shanghai International Conference Center ,pp.4074-4079, May 9-13, 2011.

[27].	Will Maddern, Michael Milford, Gordon Wyeth, " Continuous Appearance-based Trajectory SLAM," 2011 IEEE International Conference on Robotics and Automation Shanghai International Conference Center pp.3595-3600, May 9-13, 2011.
[28].	Minjie Liu, Shoudong Huang, Gamini Dissanayake, " Feature Based SLAM Using Laser Sensor Data with Maximized Information Usage, " 2011 IEEE International Conference on Robotics and Automation Shanghai International Conference Center,pp.1811-1816, May 9-13, 2011.
[29].	Hong-jian Wang, Jing Wang, Le Yu, Zhen-ye Liu, " A new SLAM method based on SVM-AEKF for AUV, " 2011 MTS.
[30].	Young Hoon Lee, ChangHyun Jun, HyungA Choi, Soo-Hyun Ryu, Nakju Lett Doh," The analysis of effect of observation models and data associations on the consistency of EKF SLAM, " The 8th International Conference on Ubiquitous Robots and Ambient Intelligence Nov. 23-26, 2011 IEEE.
[31].	Data Sheet ,Scanning Laser Range Finder ,UTM-30LX/LN,Hokuyo Automatic,2008.04.14