Abstract-This classifying pixels in an image according to

Abstract-This
paper presents the comparative analysis of texture image classification using
three methods. Texture is a repeating pattern of local variation in image
intensity. The texture provides information in the spatial arrangement of
colors or intensities in an image, thus texture is a feature used to partition
images into regions of interest and to classify those region. The
content based image retrieval technique (CBIR) is very effective if
classification of large scale general purpose image database into textured and
non textured images is done. A technique to accurately classify the images into
textured or non textured category is based on image features. In this paper we
present three methods comparison purpose for classification of textured image.
The third method is proposed method based on neural network method excepting
that gives better accuracy for image classification.

Keywords- Textured image, Support Vector Machines, Grey Level, Image
segmentation, Wavelet transforms.

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!


order now

1.Introduction-Texture
classification is important in content based image retrieval (CBIR) system. The
CBIR is technique for retrieving semantically relevant images from an image
database based on automatically derived image features. Texture classification
is concerned with identifying given textured region from given set of textured
classes. The texture classification is basically classifying pixels in an image
according to their texture cues .Three principles approaches used in image
processing to describe the texture of region are  Statistical ,Spectral, Support Vector Machines
In this paper the following three methods were discussed

1.      
Classification of
image using color and texture attributes.

2. Texture Image
Classification Using Support Vector Machine

3. Texture
Classification Based on Neural network and wavelet Transform

 

2.Classification of image using
color and texture attributes.-In this method we
propose an algorithm to improve the accuracy of this classification by
employing wavelet transform for extraction of feature for monochrome as well as
color images. We use an algorithm to
classify a Photographic image as textured and non textured, using region
segmentation and statistical testing. 1

The algorithm uses well known  LUV color space where L encodes frequency
information (luminance). U and v encodes color information (chrominance).To
obtain remaining three feature the Harr wavelet transform is applied to the L
component of the image.7 The k-means algorithm is used to cluster the feature
vectors into several classes with every class corresponding to one region in
the segmented image. The k-means algorithm is a well-known statistical
classification algorithm The k-means algorithm is used to cluster the feature
vectors into several classes with every class corresponding to one region in
the segmented image. K-mean algorithm uses pixel wise segmentation instead of
block wise segmentation.3 After applying K-means clustering algorithm we
obtain different classes. To classify images into the semantic classes textured
or non-textured, a mathematical description of how evenly a region scatters in
an image is the goodness of match between the distribution of the region and a
uniform distribution. The goodness of fit is measured by the x2 statistics.

Textured and non-textured images
are classified by thresholding the average x2 statistics for all the regions in
the image.

                                      m 

    
                 X2= 1/m ? xi2                                       2.1

                                    i=1                     

Where  Xi2 —- statistics in region i (i= 1 …m)

               X2 — average statistics 

 Where Xi2 
=

   
                2.2

                                       

If  X 2 < 0.32, the image is labeled as textured; otherwise, non-textured.  The histograms of X 2 for the two types of images are shown in Figure. It is shown that the two histograms separate significantly around the decision threshold 0.32. 3 Fig.2.1 Textured Vs Non textured image For general purpose images such as images in photo library, images on www (world wide web) etc automatic image classification for CBIR is difficult. In this method a wavelet transform based algorithm for computation of feature vector is proposed. We used a general-purpose image database containing 100 images of Dr. J. Z. Wang database. These images are pre-categorized into 10 groups: African people, beach, buildings, buses, dinosaurs, elephants, flowers, horses, mountains & glaciers, and food. All images have the size of 384x256 or 256x386. All images are stored in JPEG format.1.         Simulation Result   Sr. no Image no X12 X22 X32 X42 X52 X62 X2 Final Result   1 0 0.0448 0.0970 0.1111 0.2060 0 0 0.0765 Textured 2 4 0.1156 1.0572 0.7588 0.5204 0 0 0.4086 Non Textured 3 7 0.7242 0.4817 0.2200 0.3859 0 0 0.3020 Textured 4 17 0.8285 0.3974 1.2917 1.3316 0 0 0.6415 Non Textured 5 19 0.9563 0.3261 0.2291 0.5496 0 0 0.3435 Non Textured  Table 2.1 The Image and final result by using X2     Fig 2.2 Sample images   The image database is downloaded from website http//www.DB.Standford.edu/Image. Further statistical method is used for the classification of images into textured or non textured classes so that the search domain for the CBIR is reduced. The algorithm is compared with the standard method and found to classify the images with good accuracy From this method it is concluded that an algorithm for classification of images into textured/Non-textured images is implemented. The limitation of this method is this algorithm classify only image and to improve the segmentation and classification accuracy, the study of using the shape features into account during pixel clustering and similarity distance computation can be considered. Also it doesnot provide any information about Contrast, Correlation, Energy, and Homogeneity for texture classification .This can improve by support vector machine (SVM) technique that is discussed in second method.   3. Texture Image Classification Using Support Vector Machine- Texture is defined as a pattern that is repeated and is represented on the surface or structure of an object. To separate textures into a single texture type, first we need to preserve spatial information for each texture. For instance, the manual grey level thresholding which does not provide the spatial information for each texture that could generate in appropriate segmentation result. Grey Level Co-occurrence Probabilities (GLCP) statistics are used to preserve the spatial characteristics of a texture. The selection of certain texture is possible based on the statistical features. The best statistical features that are used for analysis are entropy, contrast, and correlation . However, further analysis in shows that correlation was not suitable for texture segmentation. GLCP statistics can also be used to discriminate between two different textures. Boundaries can be created from the shift on statistical feature while moving from one texture to another.2          Support Vector Machine (SVM) is a type of training method which is used to separate extracted features by creating a separating hyper plane . SVM have been proven to overcome the local minimum that happens in Neural Networks (NN) training algorithms. Thus, SVM provides a better performance in terms of accuracy for classification and regression. In imaging, SVM is modified to do several classification tasks, such as pattern recognition, in edge detection, in texture classification and video classification. SVM serves as complement for image segmentation methods.   3.1.1 METHODOLOGY This method considers the problem of texture classification only for a gray-level case which is conventionally tackled in two stages of feature extraction and classification.   3.1.2 GLCP Feature Extraction: GLCP is a discrete function that represents joint probability, Cij, of different sets of pixels having different grey levels, and is defined by                                           3.1   where Fij is the co-occurrence matrix constructed by the frequencies of two grey levels of two relational pixels. G represents the grey level quantization. The distance between two relational pixels is set to become 1 for micro-texture analysis. The common angle is either 0°, 45°, 90° or 135°. To reduce the computation time in GLCP feature extraction, we set a window size, M×N or a block of pixels as one feature value. 2     3.2. SVM Classification The purpose of SVM is to map feature vectors into a higher dimensional feature space, and then creating a separating hyper plane with maximum margin to group the GLCP features. Support vectors (SVs) contain highlighted pixels that help to create the margins or boundaries in an image. The higher dimensional space is defined by a kernel function. The kernel functions that we used in texture discrimination are shown in Table 3.1.     Table.3.1: Kernel functions for used in SVM training 2, 9 3.3 Simulation Result: In this SVM method the dataset is 15 types of texture images which are retrieved from bordatz database 13. Each type of texture image consists of 9 equal size samples. Texture images are rice, oriented rattan, handmade paper, fur image, pressed cork, grass, straw etc. All images are stored in PNG format. The texture image database is downloaded from website: http://perso.telecom-paristech.fr/~xia/invariant_texture/invariant_texture_brodatz/Brodatz_re.html The sample image is shown          Fig.3.1 Sample image 1_1.png Figure 3.2:The graph of GLCP statistical features generated from figure3.1 In this method the texture related parameters like contrast, correlation,enerry and homogenety are calculated and depending on these parameter the image classification is done. From above table it is shown that, for multiclass classification RBF(Radial basis function)  provide maximum correct classification rate. Therefore more multiclass classification kernel choose to be Radial basis function, it is shown in table 3.2.   Table 3.2: Experimental results for accuracy that can be achieved in Developed System 2 Sr No. Kernel parameter Classes Accuracy % training sample Time (Sec) 1 Linear 2 100 10 79.45 2   RBF with c=128,g 0.125   5 90 25 161.19 Where c = set the parameter c for regularized support vector classification,            g = set gamma in kernel function.         From table 3.2, Experimental results shows accuracy for multiclass classification by selecting kernel  RBF with c =128,g = 0.125.2 Figure 3.3 shows graph for Accuracy versus number of training samples per class. The average accuracy is achieved 80%.   Figure3.3 Accuracy of texture classification  of SVM system 2 From this method it is concluded that an algorithm for texture image classification using support vector classification is proposed and implemented. This algorithm classifies texture images using GLCP and SVM as a feature extraction and classification. SVM can be considered as a modern classification approach which features a lot of benefits, such as kernel trick and soft-margin classifiers. The drawback of this method is classification accuracy get reduced as training sample increases as well as execution time is also increased.   4. Texture Classification Based on Neural network and wavelet Transform In this method neural network and discrete wavelet transform is used for classifying textured images.7 The multi resolution analysis is applied to textured images to extract a set of intelligible features. These extracted features, in the form of DWT coefficient matrices, are used as inputs to four different multilayer perception (MLP) Neural Networks and classified. This is proposed method for texture classification. We expect that higher classification accuracy can be obtain as we increase training sample.12 Neural Network as a Classifier-The feed forward neural network, and a description of the back propagation learning algorithm is given, which is very help full for classification of texture. The basic building block of an artificial neural network is the neuron. The connection weights between neurons are adjusted. The neuron receives inputs opi from neuron  ui while the network is exposed to input pattern p. Each input is multiplied by a connection weight wij, where wij is the connection between neurons ui and uj. The connection weights correspond to the strength of the influence of each of the preceding neurons. After the inputs have been multiplied by the connection weights for input pattern p, their values are summed, net pj. Included in the summation is a bias value ?j to offset the basic level of the input to the activation function, f (net pj), which gives the output opj.12 Figure 4.1 shows the structure of the basic neuron. Fig 4.1 Basic Neuron   An artificial neural network is a system of processing elements (PE) interconnected by various synaptic strengths Recently, they have become popular classification devices for both one-dimensional and two-which use a gradient descent learning algorithm called back propagation (BP) and a topology called multilayer perceptron (MLP) have been the most dominant structure for classification purposes. Back propagation uses a squared error cost function which expresses the difference between the actual and desired responses of the network. In this method the features are extracted using Discrete Wavelet Transform (DWT) is proposed. The spatial-frequency information which a DWT contains is ideal for classifying such images as textures. The  four separable sub matrices at any given resolution level: low low(LL), LH, HL, and HH. 7 The block diagram for proposed method is as shown in fig 4.2.12   Fig.4.2 Proposed method for Texture Classification Based on Neural network and wavelet Transform   From this proposed method we are expecting better classification rate reducing execution time.   5. Conclusion From above comparative analysis it is concluded that in first method, algorithm classify only image as texture or non texture .It does not provide any information about texture. So we use second method of SVM . In this method algorithm classifies texture images using GLCP and SVM as a feature extraction But drawback of this second method is as training sample increases the classification accuracy of texture image get redued so to over come this drawback we proposed the third method in which we will try to improve the classification accuracy by using neural network and discrete wavelet transform.   References 1 Dr.Prashant.V.Ingole, Atul H.Karode, S.R.Suralkar "Textured and Non-textured image classification using wavelet transform for CBIR" National Conference on Emerging Trends in Electronics Engineering & Computing   Nagpur  ,Maharashtra state (India)  9-10 Feb. 2010. 2 S.R.Suralkar, Atul H.Karode, Ms.Priti W.Pawade "Texture Image Classification Using Support Vector Machine" International journal of Computer technique & Application (IJCTA), ISSN 2229-6093 Vol.3, issue 1 , pp 71-75, Jan -Feb  2012.  3 Jia Li, James Ze Wang, and Gio Wiederhold ,Department of Computer Science, Stanford University "Classification Of Textured And Non-Textured Images Using Region Segmentation"  IEEE transactions on Image Processing. 2000. PP   754-757 4 M. Unser, "Texture classification and segmentation using wavelet frames", IEEE transaction on Image Processing ,volume 4,Issue 11, PP 1549-1560, Nov 1995. 5 Kwang In Kim, Keechul Jung, Se Hyun Park, and Hang Joon Kim "Support Vector Machines for Texture Classification". IEEE transaction on Pattern analysis and Machine Intelligence, volume 24 Issue 11, PP 1542-1548. 6 Aditya Vailaya, Associate Member, IEEE, Mário A. T. Figueiredo, Member, IEEE, Anil K. Jain, Fellow, IEEE, and Hong-Jiang Zhang, Senior Member, IEEE "Image Classification for Content-Based Indexing". IEEE transaction on Image Processing ,volume  10, issue 1, PP 117-120, Jan2001. 7 R.C Gonzalez, R.E Woods, Steven L Eddins, Chapter 7 "Wavelets" ,PP 331-373 "Digital Image Processing Using MATLAB, Mc Graw Hill education 2010 Second  edition. 8 Rajpoot K.M."Wavelets and support vector machines for texture classification"  Multi topic Conference, 2004, Proceedings of INMIC 2004, 8th International Conference, pp. 328 – 333. 9 Kwang In Kim, Keechul Jung, and Jin Hyung Kim, "Texture-Based Approach for Text Detection in Images Using Support Vector Machines and Continuously Adaptive Mean Shift Algorithm" IEEE Transactions On Pattern Analysis And Machine Intelligence, Volume 25, No.12, December 2003. 10. Lalit Gupta, Sukhendu Das, Shivani G. Rao; "Classification of Textures in SAR Images using multi-channel multi-resolution filters"; NCIP-2005, March-2005, NIAS IISc. Bangalore, India, pp. 198-201. 11 Hong-ChoonOng, Hee-KooiKhoo, "Improved Image Texture Classification Using Grey Level Co-occurrence Probabilities with Support Vector Machines Post-Processing", European Journal of Scientific Research ISSN 1450-216X Vol.36 No.1 (2009), pp.56-64. 12 Ingole A. B. and Roopa Kakkeri " Texture Classification Based On Neural Network and wavelet Transform"  International Journal  N N A, Serial Journals, 5(1) January-June 2012, pp. 59-63 13 P. Brodatz, Textures Database. A Photographic  Album for Artists and Designers. Dover  Publications, 1966.            

x

Hi!
I'm Johnny!

Would you like to get a custom essay? How about receiving a customized one?

Check it out