Author : D.V. Lalita Parameswari 1
Date of Publication :7th March 2016
Abstract: Image clustering analysis plays an important role in data mining applications which groups set of pixels. Traditional approaches of clustering are based on deviation of the Euclidean distance which leads to the clusters of spherical shapes and input parameters are to be specified which are hard to determine. To overcome this, density based clustering techniques like DBSCAN,OPTICS are used to cluster satellite images. Thus, only low-dimensional images can be processed with limited computer memory and computing speed. This paper emphasizes on the implementation of P-Tree which requires very less memory and is very efficient for lossless image representation and compression. Thus a new Peano count tree (P-Tree) method is proposed on DBSCAN and OPTICS clustering techniques on satellite images. The DBSCAN and OPTICS clustering techniques are implemented on satellite images by applying P-tree structure. These techniques are compared and analyzed with accuracy and kappa statistic performance measures. It is ascertained that the performance of DBSCAN and OPTICS clustering techniques with P-tree is efficient than the clustering without P-tree. Further the accuracy and kappa statistic are better for OPTICS method than DBSCAN.
Reference :
-
- Amlendu Roy, William Perrizo, "Peano Count Tree Technology”, “Deriving High Confidence Rules from Spatial Data using Peano Count Trees”, LNCS 2118, July 2001.
- D.V. Lalita Parameswari, Dr. M. Seetha, Dr. K.V.N. Sunitha “ An improved grid based density methods for image clustering “ published in International Journal of Computer Engineering and Applications (IJCEA) ISSN: 2321-3469, Volume IX, Special Issue
- Ester M., Kriegel H.- P., Sander J., Xu X., A Density-Based Algorithm for Discovering Clusters in Large Spatial Databases with Noise, Proc. 2nd Int. Conf. on Knowledge Discovery and Data Mining (KDD-96), Portland, OR, 1996, pp. 226- 231, 1996.
- Introduction to data mining, Pang-ning Tan, Vipin kumar, Michael Steinbach. Jiawei Han, Micheline Kamber, “Data Mining: Concepts and Techniques”, Morgan Kaufmann, 2001
- J.O. Lapeyre and R. Strandh, “An efficient unionfind algorithm for ex-tracting the connected components of a large-sized mage,” Lab. Bordelais de Recherche en Inf., Bordeaux, France, Jan. 2004.
- Tech. Rep. Mihael Ankerst, Markus M. Breunig, Hans-Peter Kriegel, Jörg Sander.”OPTICS: Ordering Points to Identify the Clustering Structure”. Proc. ACM sigmod’99 Int. Conf. on Management of Data, Philadelphia PA, pp.49- 60,1999
- Martin Ester Hans-Peter Kriegel, Jörg Sander, and Xiaowei Xu. A Density-Based Algorithm for Discovering Clusters in Large Spatial Databases with Noise. Published in Proceedings of 2nd International Conference on Knowledge Discovery and Data Mining (KDD-96) pp.226-231, 1996.
- Matheus C.J.; Chan P.K.; and Piatetsky-Shapiro G. 1993. Systems for Knowledge Discovery in Databases, IEEE Transactions on Knowledge and Data Engineering 5(6): 903-913.
- Maleq Khan, Qin Ding, and William Perrizo, kNearest Neighbour Classification on Spatial Data Streams, 2002.
- Mohammad Hossain, Amal Shehan Perera and William Perrizo, Bayesian Classification on Spatial Data Streams Using P-Trees, Computer Science Department, North Dakota State University, Fargo, ND 58105, USA, 2002.
- Qin Ding, Maleq Khan, Amalendu Roy and William Perrizo, “The P-tree Algebra”, Computer Science Department, North Dakota State University Fargo, ND 58105-5164, USA, 2002
- Raymond T. Ng and Jiawei Han, CLARANS: A Method for Clustering Objects for Spatial Data Mining, IEEE transactions on Knowledge and Data engineering, Vol. 14, No. 5, September 2002
- Y. Tarabalka, J. A. Benediktsson, J. Chanussot, “Spectral-Spatial Classification of Hyperspectral Imagery Based on Partitional Clustering Techniques,” IEEE Trans. Geosci. Remote Sensing, vol. 47, no. 8, pp. 2973-2987, 2009.
- William Perrizo, Qin Ding, Qiang Ding, Amlendu Roy, “Deriving High Confidence Rules from Spatial Data using Peano Count Trees”, SpringerVerlag, LNCS 2118, July 2001