Handling of Homogeneous Regions for Image Cube Trajectory Analysis

Image Cube Trajectory (ICT) analysis is a new and robust method to estimate the 3D structure of a scene from a set of 2D images. The motion of points in 3D space is represented by trajectories in an image cube. The advantage of this method is that the motion information of a single 3D point can be jointly represented and analyzed for all available views of an image sequence. In this paper we will discuss the problem of 3D structure estimation in the case of homogeneous image regions. We will present a new approach for detection and compensation of depth artifacts which is based on the construction of a particular search space. Homogeneous regions in the image cube cause special patterns in this space. They can be detected by standard image processing methods. At the simple example of a linear moving camera we demonstrate the efficiency of the proposed algorithm.