资 源 简 介
应用背景用于图像的分割，边缘的检测等；We have shown that, contrary
to recent trends, it is possible to achieve excellent boundary detection results
using very local information and low-dimensional feature spaces. This is achieved
through the use of a novel statistical framework based on pointwise mutual
information.关键技术Detecting boundaries between semantically meaningful ob-
jects in visual scenes is an important component of many vision algo-
rithms. In this paper, we propose a novel method for detecting such
boundaries based on a simple underlying principle: pixels belonging to
the same object exhibit higher statistical dependencies than pixels be-
longing to di erent objects. We show how to derive an anity measure
based on this principle using pointwise mutual information, and we show
that this measure is indeed a good predictor of whether or not two pixels
reside on th
