三目測距以及雙目測距的比較與實作

Comparison and Implementation of Trinocular Vision and Stereo Vision

丁緯豪(Wei-Hao Ding)

指導教授 ： 吳章銘

中原大學/電機資訊學院/電子工程學系/碩士(2023年)

https://doi.org/10.6840/cycu202301692

摘要

距離對於現代的各行各業都有深遠的影響，像是自動駕駛在車流中對於前後車距的判斷；工業智能機器人需要知道距離去移動物體，都需要用到便宜又容易取得的感測器，而相機就是一種不錯的選擇，而使用單一相機測量距離對於複雜的環境容易誤判，刺眼的燈光對於使用雙目相機測量距離判斷距離又會導致無法量測，所以研究了三目測距來試圖解決這些問題，本論文會介紹在過去已經發表過測距方法，之後以雙目測距的基礎理論進行探討並且實作說明其優缺點，並且在其理論上提出一種使用三台相機進行測距的方法後，探討兩者的精度、實用性等等並分析其兩者的優點與缺點。

關鍵字

測距 ； 雙目 ； 三目

並列摘要

Distance has a profound impact on all walks of life in modern times, such as autonomous driving's judgment of the distance between front and rear vehicles in traffic; industrial intelligent robots need to know the distance to move objects, and they all need to use cheap and easy-to-obtain sensors. A camera is a good choice, but using a single camera to measure distance can easily lead to misjudgment in complex environments, and dazzling lights can make it impossible to measure distance using a binocular camera. Therefore, triocular distance measurement has been studied. Trying to solve these problems, this paper will introduce the ranging method that has been published in the past, then discuss the basic theory of binocular ranging and illustrate its advantages and disadvantages in practice, and theoretically propose a method using three cameras After ranging methods, discuss the accuracy, practicality, etc. of the two and analyze their advantages and disadvantages.

並列關鍵字

distance measurement ； binocular ； trinocular

參考文獻

[1] J. Zhang, S. Singh, and J. Nieto, “LOAM: Lidar Odometry and Mapping in Real-Time, ” Science and Systems Vol. 2, No. 9, pp. 1-9, 2020.

[2] A. Kumar, “Computer-Vision-Based Fabric Defect Detection: A Survey, ” IEEE Transactions on Industrial Electronics Vol. 55, No. 1, pp. 348-363, 2008.

[3] L. Fu. , J. Maand, and Y. Chen, “Automatic Detection of Lung Nodules Using 3D Deep Convolutional Neural Networks, ” J. Shanghai Jiaotong Univ. (Sci.) 24, pp. 517–523, 2019.

[4] 維基百科，自由的百科全書, “卷積神經網絡, ” 取自： https://zh.wikipedia.org/wiki/%E5%8D%B7%E7%A7%AF%E7%A5%9E%E7%BB%8F%E7%BD%91%E7%BB%9C.

[5] 維基百科，自由的百科全書, “電腦視覺, ” 16 4 2015. 取自： https://zh.wikipedia.org/wiki/%E8%AE%A1%E7%AE%97%E6%9C%BA%E8%A7%86%E8%A7%89.