Statistical Scene Analysis:Boundary Estimation of Objects in Presence of Noise.

摘要

Automatic computer determination of object boundaries has attracted many researchers in the various areas of image processing for some time. The concept of boundary determination is basically motivated by problems in textural analysis (a subject in scene analysis) and robotics (a subject in artificial intelligence). In this dissertation, an object boundary estimator, applicable to a class of noisy images, will be introduced. The images considered will contain an object of interest within the background. The boundary estimator will be designed based on the existing estimation techniques through certain available statistics of the object, background, noise, and the boundary. Extensions of the estimator to multi-object images will be discussed as well. The procedures developed will be recursive and readily implementable on a digital computer. A few examples will be considered to illustrate the performance of the estimator. (Author)