COMPARING HISTOGRAM BASED SEGMENTATION AND EXPECTATION MAXIMIZATION TECHNIQUES FOR OIL SPILL IDENTIFICATION

Abstract

This work aimed to compare histogram based segmentation and expectation maximization techniques for oil spill
detection and identification using ASAR images in Gulf of Mexico. This comparing of algorithms helps us to find out tracking
of oil spill, oil spill area, dark patches and spill patterns in radar images which help in regular monitoring of oil spill
coverage area. As we know histogram based analysis requires one pass through pixel in satellite image. Histogram is
computed according to all pixel and help to locate cluster in images. To measure image it can considered intensity and
colour for input data where as in expectation maximization techniques it is used to determine maximum likelihood parameter
of statistical model. It works as iterative model to find maximum posteriori (MAP) of parameters, which depends upon
unobserved latent variable. After analysing both algorithm it results that expectation maximization is more suitable for
detection of oil spill with less time duration using radar images than histogram based segmentation