Jyoti Singhai

A survey of cast shadow detection algorithms

Cast shadows need careful consideration in the development of robust dynamic scene analysis systems. Cast shadow detection is critical for accurate object detection in video streams, and their misclassification can cause errors in segmentation and tracking. Many algorithms for shadow detection have been proposed in the literature; however a complete, comparative evaluation of existing approaches is lacking. This paper presents a comprehensive survey of shadow detection methods, organised in a novel taxonomy based on object/environment dependency and implementation domain. In addition a comparative evaluation of representative algorithms, based on quantitative and qualitative metrics is presented to evaluate the algorithms on a benchmark suite of indoor and outdoor video sequences.

Image enhancement method for underwater, ground and satellite images using brightness preserving histogram equalization with maximum entropy

Visibility in an underwater and satellite images is poor, also light is strongly attenuated in water, producing images of low contrast and little color variation. Image preprocessing, smoothing, contrast stretching and restoration technology is concerned with producing and re-establishing an actual array of pixels for object representation to enhance the slow moving raw images. In this paper, an image processing method has been proposed for enhancing various slow motion underwater, ground, and satellite images, taken from underwater submarines and celestial sites. In the suggested method after noise smoothing & contrast stretching, image is equalized for better contrast using histogram equalization (HE), however, it tends to change the mean brightness of the image. So this paper proposes a novel extension of histogram equalization, actually histogram specification, to overcome such drawback as HE, named brightness preserving histogram equalization with maximum entropy (BPHEME), Experimental results show that BPHEME can not only enhance the image effectively, but also preserve the original brightness quite well.

Enhanced node cooperation technique for outwitting selfish nodes in an ad hoc network

In an ad hoc network, selfish nodes do not cooperate in packet forwarding of other nodes to maximise their own welfare. Such nodes will have a negative effect on the performance of the network. In this study, an enhanced node cooperation technique (ENCT) is proposed which is a hybrid technique of reputation-based mechanism and incentive-based mechanism. It evaluates relaying nodes’ behaviour, calculates its reputation and outwitting selfish nodes by utilising receipts submitted by each relaying node in the path to a centralised authority rather than using promiscuous mode. A mathematical analysis is provided which demonstrates that the expected gain of reputation for a particular node is maximum if it adopts cooperative behaviour. To motivate nodes for cooperation, variable incentives are provided proportionate to their reputation and current behaviour. Simulation results demonstrate modification of reputation and incentive with varying number of selfish nodes, traffic connections, and topology with static and dynamic behaviour of selfish nodes. The result shows that the ENCT is robust, fair and efficient technique.

Sequential minimal optimization for support vector machine with feature selection in breast cancer diagnosis

Accurate and proper diagnosis in shorter time avoids the breast cancer death. The goal is to find breast cancer as early as possible because earlier staging of breast cancer is curable. Support Vector Machine is a useful classifier among other method but the main disadvantage of Support Vector Machine (SVM) is that its time-consuming to train large dataset because of the traditional Quadratic Programming (QP) optimization problem. In this paper; Sequential Minimal Optimization (SMO) for SVM with feature selection in breast cancer diagnosis has been proposed. This method is more efficient on diagnosis that increases the classification accuracy with faster training time to train the datasets. We have done experiments on different training-test sets of the Wisconsin breast cancer dataset (WBCD) which is the most popular dataset among the researchers for breast cancer diagnosis. After that, performance evaluation is measured which shows the diagnostic performance of the SVM. At last, proposed approach obtained 100% accuracy with faster training time and there is no misclassification sample because false positive (FP) and false negative (FN) is zero for the model 4 in 80-20% training-test dataset.

Node stability based clustering algorithm for mobile ad hoc networks

Node mobility is an important factor at the time of clustering in ad hoc network because it directly affects the stability of cluster. In the proposed work a distributed clustering algorithm which uses node mobility as a main concern with battery power and connectivity in terms of no. of neighbouring nodes for selecting cluster heads to achieve better stability is being introduced. The proposed node stability based clustering algorithm (NSBCA) improves cluster stability, manageability and energy efficiency in MANET. The algorithm does not require to know the location of the node to be monitored by external means like GPS or mobility to be calculated by monitoring signal strength, the node itself monitor and records its movement and uses this information to elect cluster-head.

Probabilistic: A Fuzzy Logic-Based Distance Broadcasting Scheme For Mobile Ad Hoc Networks

An on-demand route discovery method in mobile ad hoc networks (MANET) uses simple flooding method, whereas a mobile node blindly rebroadcasts received route request (RREQ) packets until a route to a particular destination is established. Thus, this leads to broadcast storm problem. This paper presents a novel algorithm for broadcasting scheme in wireless ad hoc networks using a fuzzy logic system at each node to determine its capability to broadcast route request packets, based on the node location. Our simulation analysis shows a significant improvement in performance in terms of routing overhead, MAC collisions and end-to-end delay while still achieving a good throughput compared to the traditional AODV.

Image Registration: A Review of Elastic Registration Methods Applied to Medical Imaging

Image registration is mapping different images of the same scene which have undergone changes due to time, position or inherent changes in the image. In medical images these may be due to motion artifact, breathing, heartbeat, etc which are difficult to register only by rigid registration and need a local transformation to correct the deformations that are elastic in nature at the local level. This is corrected by applying elastic registration applied to the initially rigid registered images. Such mapping of coordinates of the image that has undergone transformation on account of one or a combination of these factors calls for hybrid registration. A number of elastic registration methods as applied to medical imaging have been discussed in this paper and results for hybrid registration of pre contrast and post contrast abdominal CT images are shown. The resulting images show increase in correlation coefficient and Mutual information after hybrid registration, a decrease in Mean square error with differences being minimised after registration. This helps in improving diagnostic accuracy. General Terms

Greedy Heuristic Based Energy Efficient Routing in Wireless Sensor Network

Most routing algorithm in wireless sensor networks uses the energy efficient path that consumes less energy. A single best path puts extra load to a specific node causing lower lifetime. If all the traffic is routed through minimum energy path, nodes of that path will depleted their battery power quickly. So instead of using single minimum cost path after some amount of transmission redirect the flow through alternate path. This paper proposes an energy efficient maximum lifetime routing algorithm. It is based on a greedy heuristic technique to maximize lifetime of the system. For achieving maximum system lifetime proposed algorithm uses the energy cost of links for constructing energy efficient path. Simulation results show that EEMLR algorithm balanced the energy for entire network as well as increases the lifetime of the network and gives the better result than AODV routing algorithm.

Curvelet transform based super-resolution using sub-pixel image registration

Super-Resolution (SR) is an approach used to restore High-Resolution (HR) image from one or more Low-Resolution (LR) images. The quality of reconstructed SR image obtained from a set of LR images depends upon the registration accuracy of LR images. However, the HR images can be reconstructed accurately by estimating sub-pixel displacement of image grid of the shifted LR image. In this paper an approach of reconstruction of SR image using a sub-pixel shift image registration and Curvelet Transform (CT) for interpolation is proposed. The curvelet transform is multiscale pyramid which provides optimally sparse representation of objects. Image interpolation is performed at the finest level in Curvelet domain. The experimental results demonstrate that Curvelet Transform performs better as compared to Stationary Wavelet Transform. Also, it is experimentally verified that the computational complexity of the SR algorithm is also reduced by using CT for interpolation.

A Multilevel Shrinkage Approach for Curvelet Denoising

This paper suggests an image restoration technique when the image is corrupted by additive white Gaussian noise. Based on the fact that the discrete curvelet transform is redundant, it proposes a scale adaptive threshold design for curvelet denoising where at each scale of curvelet transform a different threshold is applied to the transform coefficients to restore a noise free image. The strategy is to generate a set of thresholds corresponding to the various subbands of the transform whereas the traditional soft/hard thresholding applies the same threshold to each scale of transform coefficients. It is demonstrated numerically that this scheme obtains comparable performance to the state-of-the-art denoising approaches for a wide range of noise levels. Due to the adaptive support, the edges are clean and the restored images are visually pleasant.