Detection of oil seeping using k-means clustering based on color image segmentation

Detection of oil seeping using k-means clustering based on color image segmentation

Abstract

Color image segmentation has been the hotspot for the researchers in the image processing field. Color image segmentation using the k-means clustering algorithm has yielded fruitful results. An advantage resulting from the choice of color space representation could be taken to  enhance the performance of segmentation processes. The amount of information contained in the segmented objects is adopted as a measure to determine the segmentation rule. In this project the segmented object is an oil seeped area in sea surface. Here we segment all regions in a naturally captured image of sea surface and the segmented image can be used for further analysis like amount of affected area, counting affected area in large surface. Our proposed system is simulated and implemented using Matlab R2013a. Detection of oil seeping using k-means clustering based on color image segmentation

Introduction

In this paper k-means clustering technique has been proposed that has been applied to solve low-level image segmentation tasks. K-means is one of the simplest unsupervised learning algorithm that solve the well known clustering problem.  This clustering is convergent and its aim is to optimize the partitioning decisions based on a user-defined initial set of clustering that is updated after each iteration. This procedure is computationally efficient and can be applied to multidimensional data. K -means is an iterative technique that is used to partition an image into k-clusters.

The basic procedure is as follows:

• Pick k cluster centers, either randomly or based on some heuristic.
• Assign each pixel in the image to the cluster that minimizes the distance between the pixel and the cluster center.
• Re-compute the cluster center by averaging all of the pixels in the cluster.
• Repeat steps 2 and 3 until convergence is attained.

Modulate by switching between time and frequency domain

Aim and Objectives

Aim :

1.To detect the oil seeped area in the ocean around the surrounded areas.

2.Create a system that automatically segment the RGB image as it consistent

Objectives :

1.Verify the performance of our proposed work using Matlab r2013a

Proposed System :

1.Image segmentation using K-means clustering algorithm.

2.Going to process colour image segmentation

3.Regions are extracted for distinguishing other regions with oil seeped regions

Existing system

1.Only Binarization (only process on 0’s and 1’s) process is included for extracting the oil seeped area

2.Colour components in oil seeped natural images are not taken to account for segmentation process.

Drawbacks:

1.Binarization is not efficient to detect affected area.

2.They did not consider any colour component on sea image

3.So it is not applicable to all intensities in nature