2016 年 82 巻 2 号 p. 175-179
This paper presents a method of high-speed estimation for optical axes using image processing. The proposed method are planed to use in a ranging system, which estimate distances on the basis of triangulation by a monocular camera and a lamp. To define the centers of optical sources from shot images, Robust and high-speed estimation of distributions is needed because the light intensity distributions from lamps are deformed depending on the textures or other environmental factors. The optical axes are estimated on the basis of hypothesis that the diffused luminance distributions of the shot images from the camera have Gauss distributions. The proposed method uses binary codes in optical axes estimation to perform high-speed processing having robustness not affected by environmental variations.