Novel Approaches of Image Segmentation for Water Bodies Extraction Naheed Sayyed 1, Prarthana Joshi 2, Chaitali Wagh 3 Student, Electronics & Telecommunication, PGMCOE, Pune, India 1 Student, Electronics & Telecommunication, PGMCOE, Pune, India 2 Assistant Professor, Electronics & Telecommunication, PGMCOE, Pune, India 3 ABSTRACT Satellite image processing plays a crucial role for the research developments in many fields. Image segmentation is the fundamental approach of digital image processing. Images are considered as one of the most important medium of conveying information. In this work, methods like Otsu algorithm and k means algorithm is presented to extract water content. Here comparison is done between these methods to show which one is the best. I.INTRODUCTION One of the first steps in direction of understanding images is to segment them and find out different objects in them. Thus image segmentation plays a vital role towards conveying information that is represented by an image and also assists in understanding the image. Image segmentation is the process of dividing the given image into regions homogenous with respect to certain features, and which hopefully correspond to real objects in the actual scene. Segmentation plays a vital role to extract information from an image to create homogenous regions by classifying pixels into groups thus forming regions of similarity. The homogenous regions formed as a result of segmentation indwell pixels having similarity in each region according to a particular selection criteria e.g. Intensity, color etc. Segmentation plays an important role in image understanding, image analysis and image processing. Traditionally, lots of methods are utilized for the analysis and determine some resources like water which are Naheed,Prarthana,Chaitali Page 40
becoming extinct in nature. Segmentation subdivides an image into its constituent region or object. II. SCOPE OF THE PAPER The main idea of the image segmentation is to group pixels in homogeneous regions and the usual approach to do this is by common feature. Features can be represented by the space of colour, texture and gray levels, each exploring similarities between pixels of a region. Segmentation refers to the process of partitioning a digital image into multiple regions (sets of pixels). The goal of segmentation is to simplify and change the representation of an image into something that is more meaningful and easier to analyze. Image segmentation is typically used to locate objects and boundaries (lines, curves, etc.) in images. The result of image segmentation is a set of regions that collectively cover the entire image, or a set of contours extracted from the image. Water is becoming extinct in some areas while in the rest it is causing trouble like floods. The salient image regions are often useful for applications like image segmentation, adaptive compression and region-based image retrieval and obtained by mapping the pixels into various feature spaces, where upon they are subjected to various grouping algorithms. PROPOSED ALGORITHMS A. Image Segmentation Based on Improved Otsu Algorithms This algorithm is widely used because of its simple calculation and stability. Here the algorithm works on only gray value of the image. Thresholding is an important technique in image segmentation applications. The basic idea of thresholding is to select an optimal gray-level threshold value for separating objects of interest in an image from the background based on their gray-level distribution. While humans can easily differentiable an object from complex background and image thresholding is a difficult task to separate them. The gray-level histogram of an image is usually considered as efficient tools for development of image thresholding algorithms. Thresholding creates binary images from grey-level ones by turning all pixels below some threshold to zero and all pixels. Naheed,Prarthana,Chaitali Page 41
about that threshold to one. If g(x, y) is a threshold version of f(x, y) at some global threshold T, it can be defined as, g(x, y) = 1 if f(x, y) T = 0 otherwise Algorithm steps for otsu: 1. I = Input Image. 2. Obtain the histogram values (h) of the image I. 3. Set the initial Threshold value: Tin 4. Segment using Tin. This will produce two groups of pixels: C1 and C2. 5. Repeat step-3 to obtain the new threshold values for each class. (TC1 &TC2). 6. Compute the new threshold value T. 7. Repeat the steps 3-6 until the difference in T in successive iterations is not tends to zero. B. Image Segmentation Based on K means Algorithm It is one of the simplest unsupervised learning algorithms that solve the well known clustering problem and is used to cluster observations into groups of related observations without any prior knowledge of those relationships. The procedure follows a simple and easy way to classify a given data set through a certain number of clusters (assume k clusters) fixed a priori. The main idea is to define k centroids one for each cluster. These centroids should be placed in a cunning way because of different location causes different result. So, the better choice is to place them as much as possible far away from each other. There are always K clusters. There is always at least one item in each cluster. The clusters are non-hierarchical and they do not overlap. Every member of a cluster is closer to its cluster than any other cluster because closeness does not always involve the centre of clusters. Algorithm Steps for K means: Naheed,Prarthana,Chaitali Page 42
1. To classify a given data. 2. These should be placed in a cunning way. 3. Data set and associate it to the nearest centroid. 4. Continue the process until no point is pending 5. k new centroids. 6. Centroids do not move any more OBSERVATIONS As far as the experiment concerns, we have following observations: Table 2. Comparison between two methods: As shown in above table fig A is river Ganga, B is Landset and C is Bondenseee lake. Fig D,E,F are the results obtained by Otsu Thresholding. And fig G,H,I are the results of k mean clustering. It states that the objective function of Otsu method is equivalent to that of K means method in multilevel thresholding. They both are based on the same criteria that minimize the within-class variance. Also the Otsu method works on global thresholding while the K means method work on the local thresholding. The Otsu method requires computing a gray level histogram before running while K means does not require computing a gray level histogram before running. Both methods produce good segmentation result but K means give better results comparatively to Naheed,Prarthana,Chaitali Page 43
Otsu. Otsu method takes comparatively more time and increases the complexity of the algorithm. ADVANTAGES Below is the list of advantages due to the usage of the technique mentioned in this paper. Simplest to Implement and to run It works really well with large data base When variables are huge then k means is computationally faster CONCLUSION K means clustering algorithm works well and helps in extracting color of the desired information from the image while edge detection and threshold level also helps to extract fine details from the satellite image. The k means result highlights the blue nuclei which required for extracting the water bodies. This paper studies about the two techniques and concluded that k means gives better result in comparative ways. ACKNOWLEDGMENT We would like to take this opportunity to express our hearty gratitude and sincere thanks towards our guide Prof. Chaitali Wagh for his invaluable assistance that we have received, throughout the development of our project. We express our sincere thanks to the H.O.D. of E&TC Dept. Prof. Shilpa Phatalwalkar, and our respected Principal Dr. More for permitting us to present this project seminar and also to all staff members for their encouragement and suggestions during the partial fulfillment of the project. REFERENCES [1] Anuj Bala, An Improved Watershed Image Segmentation Technique using MATLAB, International Journal of Scientific & Engineering Research Volume 3, Issue 6, June-2012. [2] Venu Shah, Archana Choudhary and Kavita Tewari, River Extraction From Satellite Iamge, IJCSI International Journal of Computer Science Issues, Vol. 8, Issue 4, No 2, July 2011. Naheed,Prarthana,Chaitali Page 44
[3]Mohmed Ali HamdiǁModified Algorithm Marker Controlled Watershed Transform For Image Segmentation Based On Curvelet Thresholdǁ,Canadian Journal On Image Processing and Computer Vision Vol. 2 No. 8, Dec. 2011 [4] S.Mary Praveena, Dr.IlaVennila, Optimization Fusion Approach for Image Segmentation Using K-Means Algorithm, International Journal of Computer Applications (0975 8887)Volume 2 No.7, June 2010 [5] Rajiv Kumar Nath and Deb, "Water-Body Area Extraction from High Resolution Satellite Images-An Introduction, Review, and Comparison", International Journal of Image Processing (IJIP), Vol.3, No.6, pp.353-372,2009. [6] W. X. Kang, Q. Q. Yang, R. R. Liang, The Comparative Research on Image Segmentation Algorithms, IEEE Conference on ETCS, pp. 703-707, 2009 [7] L. Dongju and Y. Jian, Otsu method and k-means, in Hybrid Intelligent Systems, 2009. HIS 09. Ninth International Conference on, vol. 1, 2009, pp. 344 349. [8]Malik, Khan, Modified Watershed Algorithm for Segmentation of 2D Imagesǁ, Journal of Information Science & Information Technology, 6, No.3, pp. 546-552. 2009. [9] S. Arivazhagan and L. Ganesan. Texture Segmentation Using Wavelet Transform. Pattern Recognition Letter, 24(16):3197 3203, December 2003. [10] Gonzalez, Rafael C. & Woods, Richard E. (2002). Thresholding In Digital Image Processing, pp. 595 611. Pearson Education. [11] A. Gavlasov a, A. Proch azka, and M. Mudrov a. Wavelet Use for Image Classification. 15 th International Conference on Process Control, ˇStrbsk e Pleso, 2005. Naheed,Prarthana,Chaitali Page 45
Naheed,Prarthana,Chaitali Page 46