2016农业保险赔款给付:科技英语翻译1

It is assumed, in this work, that the discretization process
manifests itself in the lack of neatness of an image; it is
further assumed that a possible way to complete the information,
so as to increase the neatness, could be through the
statistical treatment of a set of images obtained under the
same conditions. Such treatment would allow one, among
other things, to extract the features common to all images
and, in principle, those important to describe the scene.
However, instead of the statistical treatment, this work relies
on the possible analogy between a digital image and a system
of many particles, so that results from statistical mechanics
leading to the macroscopic description of a system in
thermodynamic equilibrium may be used; in short, it is assumed
that the distribution of intensities, or intensity and
frequency among pixels will have a statistical behavior similar
to the distribution of energy among particles in a physical
system. In particular, the form for the distribution of particles
among the states of energy is used to obtain a distribution
that will accommodate pixels among the states of intensity so
as to enhance features of the object in the image. It should be
mentioned that there are other techniques,1,2 such as histogram
equalization, spatial filtering, frequency filtering, etc.,
which have been designed to improve specific features for
contrast, edge identification or for image smoothing, but not
for a global treatment of the image as proposed in this work.