V. Vijaya Kumar

An Efficient Block based Feature Level Image Fusion Technique using Wavelet Transform and Neural Network

fusion is a process of combining relevant information from two or more images into a single informative image. In this paper, wavelet transform is integrated with neural network, which is one of the feature extraction or detection machine learning applications. This paper has derived an efficient block based feature level wavelet transform with neural network (BFWN) model for image fusion. In the proposed BFWN model, the two fusion techniques, discrete wavelet transform (DWT) and neural network (NN) are discussed for fusing IRS-1D images using LISS III scanner about the location Hyderabad, Vishakhapatnam, Mahaboobnagar and Patancheru in India. Also QuickBird image data and Landsat 7 image data are used to perform experiments on the proposed BFWN method. The features under study are contrast visibility, spatial frequency, energy of gradient, variance and edge information. Feed forward back propagation neural network is trained and tested for classification since the learning capability of neural network makes it feasible to customize the image fusion process. The trained neural network is then used to fuse the pair of source images. The proposed BFWN model is compared with DWT alone to assess the quality of the fused image. Experimental results clearly prove that the proposed BFWN model is an efficient and feasible algorithm for image fusion.

Texture Segmentation Methods Based on Combinatorial of Morphological and Statistical Operations

In this paper we introduce a novel and simple image segmentation schemes that are based on combinations of morphological and statistical operations. Mathematical morphology is very attractive for this purpose because it efficiently deals with geometrical features like as size, shape, contrast or connectivity that can be considered as segmentation oriented features. The present paper derives equations on the basis of dilation, erosion and median or mean which finally results segmentation. The segmentation algorithms are divided into three groups based on number of operations and type of operations, used. Some of the proposed methods of segmentation are useful for edge based segmentation while the other is useful for region based segmentation. The segmentation quality is improved, by dynamically changing the combinatorial coefficients that are used in equations. The present combinatorial method is applied on Brodatz textures and a good segmentation is resulted.

Pattern based Dimensionality Reduction Model for Age Classification

The two most popular statistical methods used to measure the textural information of images are the Grey Level Cooccurrence Matrix (GLCM) and Texture Units (TU) approaches. The novelty of the present paper is, it combines TU and GLCM features by deriving a new model called “Pattern based Second order Compressed Binary (PSCB) image” to classify human age in to four groups. The proposed PSCB model reduces the given 5 x 5 grey level image into a 2 x 2 binary image, while preserving the significant features of the texture. The proposed method intelligently compressed a 5x5 window into a 2x2 window and derived TU on them. Thus the derived TU also represents a TU of a 5x5 window. The TU of the proposed PSCB model ranges from 0 to 15, thus it overcomes the previous disadvantages in evaluating TU’s.

Facial image retrieval based on local and regional features

Invention of the digital camera and also cell phones with powerful cameras with moderate and low pricing system has given the common man the privilege to capture his world in pictures anywhere, at any time, and conveniently share them with others. This has resulted the generation of volumes of images. These factors have created numerous possibilities and finally created interest among the researchers towards the design of an efficient and accurate Content Based Information Retrieval (CBIR) system. Thats why new technological advances and growth in CBIR has been unquestionably rapid during the last five years. Various face recognition methods are derived using local features, and among them the Local Binary Pattern (LBP) approach is very famous. The basic disadvantage of these methods is they completely fail in representing features derived from large or macro structures or regions, which are very much essential for faces. To address this present paper proposes a median based multi region LBP. The proposed median based multi region LBP, initially divides the facial image in to non-overlapped regions of size 5 × 5. LBP values are evaluated by dividing the region in to sub regions of size 3 × 3. The 9 sub-region LBP values are arranged in the sorted manner and the median LBP code is considered as the feature vector for the region. The present paper also proposes the minimum and maximum based regional LBP methods for efficient image retrieval. To overcome the noise and illumination effect the proposed method initially applied DOG preprocessing method with gamma correction. The proposed method is applied on FG-NET and Goggle databases for efficient facial image retrieval. The experimental results indicate the efficiency of the proposed method.

Extraction of Texture Information from Fuzzy Run Length Matrix

a precise texture classification and analysis, a run length matrix is constructed on the Local Binary pattern using fuzzy principles in the present paper. The proposed Run Length Matrix on Fuzzy LBP (RLM-FLBP) overcomes the disadvantages of the previous run length methods of texture classification that exist in the literature. LBP is a widely used tool for texture classification based on local features. The LBP does not provide greater amount of discriminate information of the local structure and it has a various other disadvantages. The main disadvantage of LBP is, that it compares the centre pixel value with its neighbors to derive the one of the three possible values {0, 1, 2}. The basic drawback of this comparison is that it is very sensitive to noise. And a major contrast between the central pixel and its surroundings are easily resulted by the slight fluctuations above or below the value of the Centre Pixel (CP) and its surroundings. To overcome this problem and to represent the missing local information effectively in the LBP, the present study introduced the concept of fuzzy logic on LBP. This overcomes the problem related to noise and contrast. The proposed method initially converts the 3×3 neighborhood in to fuzzy LBP. In the second stage the proposed method constructs the Run Length Matrix on Fuzzy LBP (RLM- FLBP). On these RLM-FLBP texture features are evaluated for a precise texture classification.

A New Algorithm for Skew Detection of Telugu Language Document based on Principle-axis Farthest Pairs Quadrilateral (PFPQ)

Skew detection and correction is one of the major preprocessing steps in the document analysis and understanding. In this paper we are proposing a new method called “Principle-axis farthest pairs Quadrilateral (PFPQ)” mainly for detecting skew in the Telugu language document and also in other Indian languages. One of the popular and classical languages of India is Telugu language. The Telugu language is spoken by more than 80 million people. The Telugu language consists of simple and complex characters attached with some extra marks known as “maatras” and “vatthulu”. This makes the process of skewing of Telugu document is more complex when compared to other languages. The PFPQ, initially performs pre-processing and divides the text in to connected components and estimates principle axis furthest pair quadrilateral then removes the small and large portions of quadrilaterals of connected components. Then by using painting and directional smearing algorithms the PFPQ estimates the skew angle and performs the de-skew. We tested extensively the proposed algorithm with five different kinds of documents collected from various categories i.e., Newspapers, Magazines, Textbooks, handwritten documents, Social media and documents of other Indian languages. The images of these documents also contain complex categories like scientific formulas, statistical tables, trigonometric functions, images, etc. and encouraging results are obtained.

Multi-Tenancy authorization system in multi cloud services

The recent days of cloud computing, the services offers measurable remarks of reducing the services costs; it helps in most of the IT infrastructure organizations. Multi-tenancy provides a single resource sharing to enhance multiple application instances. The multi-tenant feature will enable the end users to reduce the overall cost of IT infrastructure by converting many small application instances to one or few large instances. In this paper, the authors present a structural design for accomplishing multi-tenancy at the Service-Oriented Architecture (SOA) level, it allows users to route multi cloud services to use multi-tenant submissions. The contributions of this paper is design of a multi-tenant services platforms, which results the end-users can perform their tasks easily on multi-tenant environments.

Image Retrieval using Quantized Local Binary Pattern

Image retrieval is one of the main topics in the field of computer vision and pattern recognition. Local descriptors are gaining more and more recognition in recent years as these descriptors are capable enough to identify the unique features, which suitably and uniquely describe any image for recognition and retrieval. One of the popular and efficient frame works for capturing texture information precisely is the Local binary pattern (LBP). LBP descriptors perform well in varying pose, illumination and lighting conditions. LBP is a structural approach and plays significant role in wide range of applications. One of the disadvantages with LBP based frame work is its dimensionality. The dimensionality of LBP increases, if one increases the number of neighboring pixels. Further statistical approaches gained lot of significance in image retrieval and LBP based methods raises high dimensionality and complexity issues, in deriving statistical features. The present paper addresses these two issues by quantizing the LBP code, to reduce dimensionality and by deriving GLCM features on quantized LBP. The proposed method is experimented on Corel database and compared with other existing methods. The experimental results indicate the high retrieval rate by the proposed method over the existing methods.