ABSTRACT
In this chapter, the authors cover the fundamental concepts of epipolar geometry. They discuss mathematical finesse to implement the two applications from disparity and optical flow, have been greatly simplified to get the fundamentals across. Epipolar geometry defines geometric constraints across multiple cameras capturing the same scene. Fun Facts epipolar geometry seems to have been first uncovered by von Guido Hauck in 1883. He wrote several papers on the trilinear relationships of points and lines seen in three images. The authors examine some applications of epipolar geometry. In particular, when dealing with uncalibrated cameras, epipolar geometry provides some constraints which can be used to derive different geometric scene parameters like depth. The authors explore estimation of fundamental matrix for different camera pair setups. They describe a very specific type of camera setup like the two frontal eyes in animals like humans. The authors derive formally the equations they need to reconstruct depth from disparity.
