Optimisation-Based Image Grid Smoothing for SST Images - Advanced Concepts for Intelligent Vision Systems - page 226

Optimisation-Based Image Grid Smoothing for
SST Images

Guillaume Noel, Karim Djouani, and Yskandar Hamam

French South African Institute of Technology, Tshwane University of Technology,
Pretoria, South Africa

Abstract. The present paper focuses on smoothing techniques for Sea
Surface Temperature (SST) satellite images. Due to the non-uniformity
of the noise in the image as well as their relatively low spatial resolution,
automatic analysis on SST images usually gives poor results. This paper
presents a new framework to smooth and enhance the information con-
tained in the images. The gray levels in the image are filtered using a
mesh smoothing technique called SOWA while a new technique for res-
olution enhancement, named grid smoothing, is introduced and applied
to the SST images. Both techniques (SOWA and grid smoothing) repre-
sent an image using an oriented graph. In this framework, a quadratic
criterion is defined according to the gray levels (SOWA) and the spatial
coordinates of each pixel (grid smoothing) and minimised using non-
linear programming. The two-steps enhancement method is tested on
real SST images originated from Meteosat first generation satellite.

Keywords: Grid smoothing, Graph-Based approach, Non-linear opti-
misation, SST, Remote sensing.

1 Introduction

The temperature of the ocean surface reflects important underlying oceano-
graphic processes related to marine organisms and ecosystem dynamics. Areas
of special interest due to their strong biological activity, thermal fronts are nar-
row regions of separation between two large areas of homogeneous temperature
on the ocean surface. The water circulation associated with thermal fronts is
responsible for the transportation system of the ocean. Oceanographers study
these physical structures and create indices that interface the physical processes
to the biological processes from which they can study the marine ecosystem and
the marine fish population [1]. The behaviour of ocean mesoscale structures are
usually modelled by a two layers ocean model, which ensures the continuity of
the spatial derivatives of the temperature in at least one spatial direction [2].
The properties of the structures are not always depicted in the SST images due
to the noise and the low spatial resolution of the image. Sea surface temperature
(SST) images retrieved from satellites contain noise introduced by different at-
mospheric sources that complicates automatic detection. Clouds absorb infrared
emission and limit the information that is available on each SST image [3]. The