Depth computations using data in the form of intensity of pixels for enhanced computer vision

Literature review on computer vision revealed that considerable work was carried out on making the computer to understand objects and their relative positioning using techniques such as shape from shading etc. No specific references were made on quantitative estimation of depth/depth reconstruction between two points in a given 2D image of a 3D environment. An attempt is made in this paper to address above issue by proposing a second order equation using numerical methods connecting the depth between two points in the given 2D image of a 3D environment and the corresponding difference in pixel intensity values between the same two points in the 2D image. The validity of the proposed equation is verified by comparing with existing methods. The major contribution of the present work is that the depth reconstructed using the proposed equation is in excellent agreement with the corresponding depth using existing methods.