光斑測距是利用紅外線相機以及紅外線投射器所組成的3D結構光深度感應器，藉由對空間進行編碼以及由晶片解碼來獲取空間的三維信息，目前已於微軟推出的Kinect for Xbox 360以及華碩Xtion供人機介面應用。
在研究光斑測距來獲取環境深度的過程中，可發現此項技術受限於紅外線天生的某些限制，在特殊材質、反射面以及同波長的光源干擾下而無法作用。立體視覺(Stereo Vision)是電腦視覺的一種重要形式。它是基於視差原理並利用成像設備從不同的位置獲取被測物體的兩幅圖像，通過計算對應點的偏差，來取得三維信息資訊。本論文利用Kinect內建的紅外線相機以及RGB彩色相機實現跨模態(cross-modal)立體視覺，藉由兩個獨立的系統所得到的深度，彌補儀器天生之缺陷以及對深度進行校正。
本論文實現了不同種類相機間的跨模態立體視覺，並在不使用其他感應器的情形下，得到光斑測距無法測得的區域，並藉由兩個系統測得之深度相互校正，得到更精準的深度資訊，完整重建三維場景。

Light speckle distance measuring is a depth sensor composed of 3D structured light comprised of infrared rays camera and infrared rays projector. By coding the space and decoding via the chip, we can get the depth information of the space. It has been applied on the use of Kinect for Xbox 360 of Microsoft and Xtion of ASUS for man-machine interface.
During studying of light speckle distance measuring, we find the technique is limited by some restrictions of infrared rays, such as special material, reflective surface and the same wavelength. Stereo vision is an important form of computer vision. It is based on disparity and utilizes imaging device to get two images in different position by calculating the deviation to get the 3D information.
In this paper, we develop the cross-modal stereo vision using the RGB camera and infrared rays camera by Kinect. Via two independence depth systems, we can remedy defects of the equipments and correct the depth.
This paper presents the cross-modal stereo vision in different kinds of cameras. Without using other sensors, we obtain the depth area which is missing from light speckle distance measuring and higher accuracy depth information by rectification of two systems and completely reconstruct the 3D scene.

[1] D. Marr , “Vision: A computational investigation into the human representation and processing of visual information” , 1981.
[2] J. H. Jean, T. P. Wu, J. H. Lai, and Y. C. Huang, “A Visual Servo System for Object Tracking Applications of Mobile Robots Based on Shape Features,” Proceedings of 2005 CACS Automatic Control Conference Tainan, Taiwan, Nov 18-19, 2005.
[3] A. Lipton, H. Fujiyoshi and R. Patil, “Moving Target Classification and Tracking from Real-time Video, ” In Proc. IEEE Workshop on Applications of Computer Vision, Princeton, NJ, 1998.
[4] K. P. Horn, B. G. Schunck, “Determining Optical Flow,” Artifical Intelligence, Vo1.17, pp. 185-203, 1981.
[5] R. Y. Tsai, “A Versatile Camera Calibration Technique for High-Accuracy 3D Machine Vision Metrology Using Off-the-shelf TV Cameras and Lenses,” IEEE Journal of Robotics and Automation, Vol.3, pp. 323-344, 1987.
[6] H. Hirschmuller, “Stereo Processing by Semi-Global Matching and Mutual Information”,IEEE Journal of Pattern Analysis and Machine Intelligence,Vol. 30, pp. 328-341, 2007.
[7] R. Y. Tsai, , “An efficient and accurate camera calibration technique for 3D machine vision,” Proc. International Conference on Computer Vision an Pattern Recognition, pp. 364-374, 1986.
[8] Z. Zhang, “A Flexible New Technique for Camera Calibration,” 1998.
[9] M. Pollefeys, Self-Calibration and Metric 3D Reconstruction from Uncalibrated Image Sequences, Ph.D. thesis, ESATPSI, K. U. Leuven, 1999.
[10] W. C. Chiu, U. Blanke, M. Fritz, “I spy with my little eye: Learning Optimal Filtering for Cross-Modal Stereo under Projected Patterns,” IEEE International Conference on Computer Vision Workshops, pp. 1209-1214, 2011.
[11] O. Faugeras, Q. T. Luong, “The fundamental matrix : Theory Algorithms, and Stability Analysis＂, Intel J. Computer Vision, 17, pp. 43-45, 1996.
[12] H. C. Longuet-Higgins, “A Computer Algorithm for Reconstruction A Scene from Two Projections＂, Nature, 293, pp. 133-135, 1981
[13] R. Hartley, “ In Defense of The Eight-Points Algorithm,＂ IEEE Trans. Pattern Analysis and Machine Intelligence, Vol. 19, pp. 580-593, 1997.
[14] T. S. Huang, O. Faugeras, “Some Properties of The E Matrix in Two-View Motion Estimation＂, IEEE Trans. Pattern Analysis and Machine Intelligence, pp. 1310-1312, 1989.
[15] W. C. Chiu, U. Blanke, M. Fritz, “Improving the Kinect by Cross-Modal Stereo,” Conference on British Machine Vision, 2011.
[16] P. F. Felzenszwalb, D. P. Huttenlocher, “Efficient belief propagation for early vision,” IJCV, 70, pp. 41 – 54, 2006.
[17] M. Fritz, M. Black, G. Bradski, S. Karayev, T. Darrell, “An additive latent feature model for transparent object recognition,” NIPS, 2009.
[18] H. Hirschmuller, D. Scharstein, “Evaluation of stereo matching costs on images with radiometric differences,” TPAMI, 31, pp. 1582 – 1599, 2009.
[19] Y. Li, D. P. Huttenlocher, “Learning for stereo vision using the structured support vector machine,” CVPR, 2008.
[20] S. Gould, P. Baumstarck, M. Quigley, D. Koller, “Integrating Visual and Range Data for Robotic Object Detection,” ECCV, 2008.
[21] G. Bradski, “The OpenCV Library,” Dr. Dobb’s Journal of Software Tools, 2000.
[22] M. Grant, S. Boyd, “CVX: Matlab software for disciplined convex programming,” http://cvxr.com/cvx, 2011.
[23] http://www.javaforge.com/wiki/101886
[24]http://www.cnblogs.com/TravelingLight/archive/2011/10/11/2207275.html
[25] http://book.2cto.com/201211/9224.html
[26] http://book.2cto.com/201211/9227.html