Dheeraj Agrawal

A meta-analysis on content based image retrieval system

The objective of this paper is to analyze the research work in the field of content based image retrieval (CBIR). Content based image retrieval is very important filed for efficient image retrieval system. This paper focuses on the latest trends available and the methodology in the current research. Based on the study several result comparison and research gaps have been discussed. The discussions provided in this paper provide a basis for the future research and also help in identifying the gaps or the area where focus can be centered in the future.

Comparative analysis of off-line signature recognition

Biometrics (or biometric authentication) assigns to the confirmation of humans by their biological features. Computer science, biometrics to be specific, is used as an aspect of determination and access control. It is also used to determine individuals in groups that are under examination. Handwriting is one of the most widely used biometric systems for authentication of person as well as document. Online and offline signature is existing in person identification and authentication problems. Offline signature categorizes the signature into two classes: genuine and forged. In this paper, we discuss various features of offline signature recognition and verification process. We review and compare existing techniques, their results and methods of feature extraction.

A Novel Approach for Diagnosis of Noisy Component in Rolling Bearing Using Improved Empirical Mode Decomposition

The Bearing is utilized to give free direct development to the moving part or with the expectation of complimentary revolution around a fixed axis. Bearings are considered a main part in various mechanical systems. Multi component vibration signals are generated when the machine works. Accelerometers are used to capture generated vibration signal. Vibration signal analysis is effectively used to diagnose bearing faults. There are various methods using empirical mode decomposition (EMD) as their fundamental method to diagnose bearing faults. The proposed method consists of analyzing the kurtosis of residue obtained after removing higher frequency components of the original signal. The proposed technique identifies the boisterous frequency segment in the signal through the iterative procedure. The experimental data were collected from Case Western Reserve University, Ohio. The simulation is done over MATLAB 7.8.1.

Comparison between empirical and variational mode decomposition based on percentage variation in entropy feature from glaucoma image

Glaucoma is a type of eye disease; it damages the optic nerve due to gradual increase in the fluid pressure and hence causes blindness. In the paper two decomposition techniques, namely, bi-dimensional empirical mode decomposition (BDEMD) and two dimensional variational mode decomposition (2DvMD) are used and compared to find the better decomposition technique. Images are decomposed by these methods and entropy features are extracted from decomposed sub band images. The percentage variations in entropy features have been calculated from the extracted features for each decomposition technique for normal and glaucoma image. These calculated percentage variations in entropy features are used to compare the two decomposition techniques for normal and glaucoma images. The results obtained put forward that the percentage variation in entropy feature extracted from 2DvMD are higher than BDEMD. Hence, 2DvMD outperforms over BDEMD.

Computer aided diagnosis of glaucoma using discrete and empirical wavelet transform from fundus images

Glaucoma is a class of eye disorder; it causes progressive deterioration of optic nerve fibres. Discrete wavelet transforms (DWTs) and empirical wavelet transforms (EWTs) are widely used methods in the literature for feature extraction using image decomposition. However, to increase the accuracy for measuring features of images a hybrid and concatenation approach has been presented in the proposed research work. DWT decomposes images into approximate and detail coefficients and EWT decomposes images into its sub band images. The concatenation approach employs the combination of all features obtained using DWT and EWT and their combination. Extracted features from each of DWT, EWT, DWTEWT and EWTDWT are concatenated. Concatenated features are normalised, ranked and fed to singular value decomposition to find robust features. Fourteen robust features are used by support vector machine classifier. The obtained accuracy, sensitivity and specificity are 83.57, 86.40 and 80.80%, respectively, for tenfold cross validation which outperforms the existing methods of glaucoma detection.

An analysis based on local binary pattern (LBP) and color moment (CM) for efficient image retrieval

This paper objective is to analyze LBP and CM for the comparative analysis in the field of content based image retrieval (CBIR). In this paper LBP and CM methods are used for the efficient retrieval of the images from the Wang database. All the 10 classes are used for the results comparison. LBP is used for better thresholding and CM is used with color dominant features for the efficient retrieval. The results suggested that the results may vary with different input images and the chances of the prediction are also different.

Vibration signature analysis using variable Tukey window: A case study on Bearing Fault Data

The bearing is used to provide free linear movement to the moving part or for free rotation around a fixed axis. When the machine is in use, the multi component vibration signals are generated. These vibration signals can be captured by accelerometers. Vibration signal analysis is a very effective tool in finding bearing fault. The methods based on empirical mode decomposition (EMD) have been used for bearing fault diagnosis in mechanical systems. Boundary distortion is considered as a serious problem in the application of EMD. High values of Kurtosis analysis of the IMFs due to boundary distortion may mislead to faults, even if the fault is not present. In EMD process, the boundary distortion of successive IMFs increases, a variable Tukey window is used in this paper to address the increasing boundary distortion problem. In the analyze method boundary distortion problem is minimized by using a variable window for all IMFs.