Abstract—In this paper, the geometry of uncertainty in depth measurement of stereo vision system is analyzed and visualized. A CCD sensor of the camera has a minimum unit called a pixel. Based on the pinhole camera, the camera's pixels pass through the lens and contain the most noticeable information among objects in a rectangle that increases in proportion to the distance. Also, in a stereo vision system composed of two or more cameras, the vertices of the pixels of each camera pass through the lens of the camera and contact each other to form a polygon. So, we call this polygon a uncertainty area, because human beings can’t determine the exact position when a noticeable object moves within this polygon. Therefore, the core of this research is to determine the shape and size of the uncertainty area in the depth measurement of the stereo vision system made up of cameras, and the ultimate goal of this research is to obtain uncertainty information in the depth measurement of the stereo vision system and apply it to sensor fusion. 
 

Index Terms—localization, robot, sensor fusion, camera, stereo vision system, stereo vision, depth measurement, uncertainty, uncertainty area, uncertainty analysis

Cite: Jinwoo Choo and Jihong Lee, "Uncertainty Analysis in Depth Measurement of Stereo Vision System for Sensor Fusion," International Journal of Electronics and Electrical Engineering, Vol. 6, No. 3, pp. 32-36, September 2018. doi: 10.18178/ijeee.6.3.32-36