Unsupervised segmentation of images with law depth of field (DOF) is highly
useful in various applications including image enhancement for digital cam
eras, target recognition, image indexing for content-based retrieval, and 3
D microscopic image analysis. This paper describes a novel multiresolution
image segmentation algorithm for low DOF images. The algorithm is designed
to separate a sharply focused object-of-interest from other foreground or b
ackground objects. The algorithm is fully automatic in that all parameters
are image independent. A multiscale approach based on high frequency wavele
t coefficients and their statistics is used to perform context-dependent cl
assification of individual blocks of the image. Unlike other edge-based app
roaches, our algorithm does not rely on the process of connecting object bo
undaries. The algorithm has achieved high accuracy when tested on more than
100 low DOF images. many with inhomogeneous foreground or background distr
actions. Compared with the state of the art algorithms, this new algorithm
provides better accuracy at higher speed.