Uma Shanker Tiwary

SOFT-THRESHOLDING FOR DENOISING OF MEDICAL IMAGES — A MULTIRESOLUTION APPROACH

Medical images generally have low contrast and they get complex type of noise due to the use of various devices and applications of various algorithms. However, most of the denoising methods consider only additive noise or some special noise model dependent on their systems and conditions only. Such methods when applied to real medical images yield poor results. The present work proposes a method for denoising of medical images using soft-thresholding in wavelet domain on multiple levels. We have developed a method to compute the threshold values for denoising of medical images, which depend on the median as well as the contrast ratio of the wavelet coefficients and also on the level number. We have performed experiments by adding various proportions of Gaussian, Salt-and-Pepper and Speckle noise, and found that the proposed method performs better for these cases. The method is efficient because the threshold values can be calculated directly and it is adaptive as these values depend on mean, median and standard deviation of wavelet coefficients of the particular image. The proposed method also gives a criterion for level-dependent thresholding. Application of the proposed method to Ultrasound, X-ray and MRI images is demonstrated in experiments. In the present work, we have also done studies on how to select the mother wavelet for a particular problem.

Multilevel adaptive thresholding and shrinkage technique for denoising using Daubechies complex wavelet transform

Abstract In this paper, we have proposed a multilevel soft thresholding technique for noise removal in Daubechies complex wavelet transform domain. Two useful properties of Daubechies complex wavelet transform, approximate shift invariance and strong edge representation, have been explored. Most of the uncorrelated noise gets removed by shrinking complex wavelet coefficients at the lowest level, while correlated noise gets removed by only a fraction at lower levels, so we used multilevel thresholding and shrinkage on complex wavelet coefficients. The proposed method firstly detects strong edges using imaginary components of complex coefficients and then applies multilevel thresholding and shrinkage on complex wavelet coefficients in the wavelet domain at non-edge points. The proposed threshold depends on the variance of wavelet coefficients, the mean and the median of absolute wavelet coefficients at a particular level. Dependence of these parameters makes this method adaptive in nature. Results obtained for one-dimensional signals and two-dimensional images demonstrate an improved denoising performance over other related methods available in the literature.

Daubechies Complex Wavelet Transform Based Moving Object Tracking

This paper describes a new method for moving object tracking, using complex wavelet transform. Real-valued wavelet transform is widely used in tracking applications, but it suffers from shift-sensitivity. Daubechies complex wavelet transform is more suitable for tracking due to approximate shift-invariance nature. The proposed method is intelligent enough to segment the object from a scene. Segmentation in the first frame has been done by computing multiscale correlation of imaginary component of complex wavelet coefficients and then object is tracked in next frames by computing the energy of complex wavelet coefficients corresponding to the object area and matching this energy to that of the neighborhood area. The proposed method is simple and does not require any other parameter except complex wavelet coefficients for segmentation as well as tracking

DAUBECHIES COMPLEX WAVELET TRANSFORM BASED MULTILEVEL SHRINKAGE FOR DEBLURRING OF MEDICAL IMAGES IN PRESENCE OF NOISE

Deblurring in the presence of noise is a hard problem, especially in ultrasonic and CT images. In this paper, we propose a new method of image deblurring in presence of noise, using symmetric Daubechies complex wavelet transform. The proposed method is based on shrinkage of multilevel Daubechies complex wavelet coefficients, and is adaptive as it uses shrinkage function based on the variance of wavelet coefficients as well as the mean and the median of absolute wavelet coefficients at a particular level. The results obtained after the application of the proposed method demonstrate an improved performance over other related methods available in literature.

Multilevel Medical Image Fusion using Segmented Image by Level Set Evolution with Region Competition

In this paper, a region level based image fusion technique, using wavelet transform, has been implemented and analyzed. The proposed methodology considers regions as the basic feature for representing images and uses region properties for extracting the information from them. A segmentation algorithm is proposed for extracting the regions in an effective way for fusing the images. The fusion strategy uses multi-level decomposition of the images obtained using wavelet transform. By analyzing the images at multiple levels, the proposed method is able to extract finer details from them and in turn improves the quality of the fused image. The performance and relative importance of the proposed methodology is investigated using the mutual information criteria. Experimental results show that the proposed method improves the quality of the fused image significantly for both the normal and multifocused images

An adaptively accelerated Lucy-Richardson method for image deblurring

We present an adaptively accelerated Lucy-Richardson (AALR) method for the restoration of an image from its blurred and noisy version. The conventional Lucy-Richardson (LR) method is nonlinear and therefore its convergence is very slow. We present a novel method to accelerate the existing LR method by using an exponent on the correction ratio of LR. This exponent is computed adaptively in each iteration, using first-order derivatives of the deblurred image from previous two iterations. Upon using this exponent, the AALR improves speed at the first stages and ensures stability at later stages of iteration. An expression for the estimation of the acceleration step size in AALR method is derived. The superresolution and noise amplification characteristics of the proposed method are investigated analytically. Our proposed AALR method shows better results in terms of low root mean square error (RMSE) and higher signal-to-noise ratio (SNR), in approximately 43% fewer iterations than those required for LR method. Moreover, AALR method followed by wavelet-domain denoising yields a better result than the recently published state-of-the-art methods.

Cognitive Model - Based Emotion Recognition From Facial Expressions For Live Human Computer Interaction

Human emotions are deeply intertwined with cognition. Emotions direct cognitive processes and processing strategies of humans. The goal of this work is to design a model with the capability of classifying the uncertainty, contradiction and the cognitive nature of the emotions. For achieving this, 3D cognitive model is designed. This model enhances our vision of classification of emotions produced by reinforcing stimuli. In this model the dimensions represent the positive reinforcers, the negative reinforcers and the emotion content present. The positive reinforcer increases the probability of emission of a response on which it is contingent, whereas the negative reinforcer increases the probability of emission of a response that causes the reinforcer to be omitted. This model increases the number of emotions, that can be classified. Presently this model can classify 22 emotions subject to the presence of a facial expression database. It has the flexibility to increase upon the number of emotions. For emotion (pattern) identification, the pose and illumination factor are removed using Gabor wavelet transforms and the size is reduced by finding its principle components (PCA). This component vector is used for training the neural network. The test result shows the recognition accuracy of 85.7% on The Cohn-Kanade Action Unit Coded Facial Expression Database. The real time processing for identification, aids in applying emotions to real time audio player. An environment, that is all pervasive or ubiquitous, that would sense ones mental state and play the appropriate musical track to maintain the positive emotional state or ease from a negative emotional state

Base station initiated dynamic routing protocol for Heterogeneous Wireless Sensor Network using clustering

Wireless sensor networks have received excessive attention in now a days. Research in all field in wireless sensor network (WSN) till now has assumed that sensors in network is homogeneous, means that all the nodes are same, especially for routing protocols that has used clustering like LEACH, LEACHC, and PEGASIS etc. But in the clustering algorithm where some nodes have to work as cluster heads, if they are same as the other nodes, this would lead the lifetime constraint in WSN. Because of this most we shall be considering here the heterogeneous WSN, where the nodes which will work as cluster head will contain more energy, computational and communication power than normal nodes. In this paper we introduce a routing protocol that is based on clustering and uses heterogeneity in nodes to increase the network lifetime.

Noniterative method for the synthesis of convergent pierce electron guns

A simple noniterative method has been proposed to design a convergent Pierce electron gun. The logarithmic value of the ratio of cathode to anode radii of curvature is expressed in a power series directly in terms of beam convergence ratio. This made it possible to calculate the half-beam cone angle with the help of Langmuir-Blodgetts solution, which, in turn, facilitated the calculation of other output parameters. The throw of the gun is also expressed in a power series in terms of the ratio of the anode-aperature radius to the beam waist radius. It is numerically shown with practical examples that both the interelectrode distance and the throw decrease with the gun perveance and increase (the former almost linearly while the latter almost binomially) with the beam diameter.

IP-based ubiquitous healthcare system

This paper presents a new concept of MAC and LOAD protocols for IP based ubiquitous healthcare system. The system used IEEE 802.15.4 standard lowpan with integrated IPv6. For healthcare system we added LOAD (6lowpan Ad-hoc on Demand Distance Vector) and MAC (Medium Access Control) protocols in Harvanpsilas 6lowpan stack. 6lowpan stack has ability to connect the physical environment in real-world applications such as healthcare, wireless sensor network, network technology etc. IP-enable motes set on the patient body for retrieving biomedical from body in PAN. PAN network connected PC via gateway or base station for further analysis or to the doctorpsilas PDA (personal digital assistant). The doctor can recognize or analysis patient data from anywhere on globe by internet service provider equipments (PDA). Result shows the performance biomedical data packets in multi-hope routing as well as represents the topology of the networks. TelosB motes were tested on octopus simulator in tinyOS2.02 for performance of biomedical data communication and network topology.