Visual experience acquisition based on view angle estimation from 3D monocular image

It has been proved that acquired training is important to the development of stereopsis experience. Month-old babies already have the initial experience of invariance recognition of 3D objects. There is a slight lack of precision in the interpretation of biological vision. However, the small cost and the fast speed in calculation meet the requirements of invariance recognition, the rich visual experience in which play an important role. But what is the experience, how to acquire and how to use, these problems have never been satisfactorily resolved. In this paper we simulate the learning of visual experience in children, and solve a view angle estimated problem by using self-organizing network, which make the hidden experience clarified. Compared to the Classic camera calibration, which a large number of parameters need to be estimated, this method needs only one image and does not aim to 3D reconstruction. By avoiding the complex calibration and registration process, an amount of computation has been reduced. Visual experiences are all obtained from the most ordinary examples, and the characterization based on the geometric feature. Therefore, this method has strong expansibility and good generalization ability.