View of IMAGE CLASSIFICATION BASED HIGH RESOLUTION SATELLITE IMAGES: A STUDY

(1)

ACCENT JOURNAL OF ECONOMICS ECOLOGY & ENGINEERING Available Online: www.ajeee.co.in Vol. 01, Issue 04, August 2016, ISSN -2456-1037 UGC APPROVED NO. 48767 (INTERNATIONAL JOURNAL)

1

IMAGE CLASSIFICATION BASED HIGH RESOLUTION SATELLITE IMAGES: A STUDY Raj Tiwari

Assistant Professor, GGITS, Jabalpur

Abstract:- A huge load of assessment has been endeavored and is being finished for developing an exact classifier for extraction of things with changing accomplishment rates.

Most of the customarily used advanced classifiers rely upon neural association or support vector machines, which uses winding reason limits, for describing the restrictions of the classes. The disadvantage of such classifiers is that the constraints of the classes as taken by extended reason work which are roundabout while the same isn't legitimate for a lot of the certified data. The restrictions of the classes vary alive and well, thusly provoking vulnerable precision.

Keywords:- Accuracy assessment, Image Segmentation, Image classification.

1. INTRODUCTION

Article based picture course of action procedures are dynamically used for request of land cover/use units from significant standard pictures, and much of the time the inevitable result is close to the way in which a human master would unravel the image. Thing based picture request doesn't work clearly on single pixels, yet picture objects which insinuate homogeneous, spatially connecting zones.

These are obtained by detaching an image, specifically picture division, which is a troublesome issue as a result of how it isn't, now important to do this task on a pixel-by-pixel premise.

The fine spatial resolution implies that each object is now an aggregation of a number of pixels in close spatial proximity, and accurate classification requires that this aspect be considered.

To deal with the problem of complexity of high resolution images, the image is first segmented into homogeneous regions, and a set of features are computed for each region segment. These segments are classified using one or more of the machine learning algorithms. In the present study, various activation functions for artificial neural network classification are considered. This method basically includes three steps.

1. Image segmentation to extract the regions from the pixel information based on homogeneity criteria.

2. Calculation of spectral parameters like mean vector, texture, NDVI and spatial/shape parameters like aspect ratio, convexity, solidity, roundness and orientation for each region.

3. Classification of image using the region feature vectors using suitable classifiers such as NN.

Supervised classification is one of the most commonly undertaken analyses of remotely sensed data.

The output of a supervised classification is effectively a thematic map that provides a snapshot representation of the spatial distribution of a particular theme of interest such as land cover. The goal of a supervised image classification system is to group images into semantic categories giving thus the opportunity of fast and accurate image search. To achieve this goal, these applications should be able to group a wide variety of unlabelled images by using both the information provided by unlabelled query image as well as the learning databases containing different kind of images labeled by human observers.

Further, these premise capacities are intended to have various limit portions, instead of a solitary limit concerning RBFs. This new upgrade to the premise work alongside a reasonable preparing calculation permits the neural organization to all the more likely become familiar with the particular properties of the difficult area.

The limits of classes considered are not round but rather a bunch of limits is considered for each class, which guarantees higher precision hypothetically. This method is underscored explicitly for multi-ghastly satellite pictures. Accordingly it was intended to propose an appropriate classifier for the high goal satellite distantly detected pictures and to test the pertinence of adjusted cloud premise capacities for the field of far off detecting.

(2)

2 2. METHODOLOGY

All things considered, a coordinated picture game plan requires three principal propels: pre-dealing with, incorporate extraction and request [1]. Considering this designing, many picture gathering structures have been proposed, each one perceived from others by the method used to deal with the image signature just as the decision procedure used in the portrayal step. Counterfeit Neural Network (ANN) and Support Vector Machine (SVM) are by and large used advanced procedures for controlled gathering of remotely recognized data [2].

The authentic detriment of SVM is that the constraints of the classes as taken by extended reason work networks are roundabout while the same isn't legitimate for prevailing piece of the veritable data. The constraints of the classes change alive and well, thusly inciting vulnerable accuracy. This work is made on the modified RBFs neural association based classifier for object based gathering of significant standard satellite indirectly distinguished pictures.

The new reason limits, called cloud premise limits use another component weighting, construed to underscore features relevant to class isolation as discussed in [3].

2.1 Proposed Methodology for Object Based Image Segmentation

The augmentation of high-spatial objective multispectral imagery from satellite and aeronautical sensors (for instance IKONOS from Geo Eye, Inc., Quick Bird from Digital Globe, Inc., ADS40 from Leica Geo systems, Inc.) has basically changed the level of refinement required in cutting edge picture planning [4]. In this paper we propose a technique for improving the accuracy of thing based controlled picture game plan using Cloud Basis Functions Neural Network for significant standard indirectly distinguished multi-spooky satellite pictures, as IKONOS or Quick Bird.

Item based picture investigation partitions the picture into significant homogeneous areas dependent on ghastly properties as well as on shape, surface, size, and other topological highlights, and coordinates them progressively as picture objects (additionally alluded to as picture fragments) [6]. The division method (extraction of the picture objects) is

constrained by the client indicated scale (size) or goal of the normal articles. Article based methodologies have been effective for land-use and land-cover characterization [7][8].

Classification of high-resolution satellite images using standard per-pixel approaches is difficult because of the high volume of data, as well as high spatial variability within the objects. One way to deal with this problem is to reduce the image complexity by dividing it into homogenous segments prior to classification. This has the added advantage that segments can not only be classified on basis of spectral information but on a host of other features such as neighborhood, size, texture and so forth.

The proposed algorithm is given below 1. Apply multi-resolution framework

(here Daubech6 family of wavelet transform is used) to input image.

2. Use multi-scale gradient algorithms to calculate color gradient.

The morphological gradient of each band of the image is calculated using equation (1)

Where

1. G(f) = Morphological color gradient,

2. f = Given image

3. B = Structuring element.

2.2 Multi Layer Feed Forward Neural Networks

The augmentation of high-spatial objective multispectral imagery from satellite and aeronautical sensors (for instance IKONOS from Geo Eye, Inc., Quick Bird from Digital Globe, Inc., ADS40 from Leica Geo systems, Inc.) has basically changed the level of refinement required in cutting edge picture getting ready [4]. In this paper we propose a technique for improving the exactness of thing based managed picture course of action using Cloud Basis Functions Neural Network for significant standard indirectly recognized multi-spooky satellite pictures, as IKONOS or Quick Bird.

Item based picture examination partitions the picture into significant homogeneous districts dependent on unearthly properties as well as on shape, surface, size, and other topological highlights, and coordinates them

(3)

3 progressively as picture objects (likewise alluded to as picture fragments) [6]. The division technique (extraction of the picture objects) is constrained by the client determined scale (size) or goal of the normal items. Item based methodologies have been effective for land-use and land- cover characterization [7][8].

3. CONCLUSION

This paper tries to study and take a gander at the precision of thing based picture classifiers. The thing based picture assessment exceptionally reduced the salt-and-pepper gathering sway in the masterminded picture without horribly impacting the portrayed picture precision.

This unfathomably improves the exceptional representation of the masterminded picture. ANN has the advantages basically of more protection from uproar wellsprings of information and depiction of Boolean limit isolated from others, yet an unnecessary number of attributes may result in over fitting.

REFERENCES

1. R.O. Duda, P.E. Hart P.E. and D.G. Stork,

“Pattern Classification”. 2nd edition, John Wiley & Sons (2001).

2. B. Tso and P.M. Mather, “Classification methods for remotely rensed data” London:

Taylor and Francis (2001).

3. C.R. De Silva, S. Ranganath and L.C. De Silva, “Cloud Basis Function Neural Network: A modified RBF network architecture for holistic for holistic Facial Expression Recognition”. Elsevier Pattern Recognition 41:1241-1253, 2008.

4. J. Im, J.R. Jensen and J.A.Tullis, “Object- based change detection using correlation image analysis and image segmentation”.

International Journal of Remote Sensing, 29(2), 399–423 (2008).

5. G.J Hay, G. Castilla, M. Wulder, and J.R Ruiz, “ An automated object-based approach for the multi-scale image segmentation of forest scenes”

International Journal of Applied Earth Observation and Geo information 7, 339–

359, 2005.

6. T. Blaschke, “A framework for change detection based on image objects”, In Go¨ttinger Geo graphics Abhandlungen, S.

Erasmi, B. Cyffka and M. Kappas (Eds), 113, 1–9, Go¨ ttingen, 2005. Available

online at

http://www.definieris.com/pdf/publication s/GGRS2004_Blaschke_G001.pdf

(accessed 18 June 2009).

7. A.S. Laliberte, A. Rango, K.M. Havstad, J.F.

Paris, R.F. Beck and Y.R. McNeely, “Object- oriented image analysis for mapping shrub encroachment from 1937 to 2003 in southern New Mexico”. Remote Sensing of Environment, 93, 198–210, 2004.

8. R.C. Frohn, K.M. Hinkel, and W.R. Eisner,

“Satellite remote sensing classification of thaw lakes and drained thaw lake basins on the North Slope of Alaska”. Remote Sensing of Environment, 97, 116–126, 2005.

9. J. Eo and H. Kim, “A Detail Extraction Technique For Image Coding Using morphological Laplacian Operator”. In Proceedings of the IEEE TENCON, pp.140- 147, Korea, 1997.

10. F. Meyer and S. Beucher, “Morphological Segmentation”. Journal of Visual Communication and Image Representation, 11: 21–46, 1990.

11. L. Costa and R. Cesar, “Shape Classification and Analysis”, 2nd edition CRC Press, 259- 310, (2009).

12. P. M. Mather, “Computer Processing of Remotely Sensed Images,” John Wiley, NY, (2004).

13. J. Han, S. Lee, K. Chi, K. Ryu,

“Comparison of Neuro-Fuzzy, Neural Network, and Maximum Likelihood Classifiers for Land Cover Classification using IKONOS Multispectral Data”. In Proceedings of the IEEE International Geo science and Remote Sensing Symposium, vol 6, pp. 3471-3473, Toronto, Canada, 2002.

14. S. Haykin, “Neural Networks: A Comprehensive Foundation” 2 nd edition, Prentice Hall pp. 156- 256, (1999)

15. T. Cover, “Geometrical and Statistical Properties of Systems of Linear Inequalities with Applications in Pattern Recognition”

IEEE Transactions on Electronic Computers, vol. EC-14(3), 326-334, June 1965. Reprinted in Artificial Neural Networks: Concepts and Theory, IEEE Computer Society Press, Los Alamitos, Calif., 1992, eds. P. Mehra and B. Wah.