Samar Bhattacharya

Plant leaf recognition using texture and shape features with neural classifiers

This paper proposes a novel methodology of characterizing and recognizing plant leaves using a combination of texture and shape features. Texture of the leaf is modeled using Gabor filter and gray level co-occurrence matrix (GLCM) while shape of the leaf is captured using a set of curvelet transform coefficients together with invariant moments. Since these features are in general sensitive to the orientation and scaling of the leaf image, a pre-processing stage prior to feature extraction is applied to make corrections for varying translation, rotation and scaling factors. Efficacy of the proposed methods is studied by using two neural classifiers: a neuro-fuzzy controller (NFC) and a feed-forward back-propagation multi-layered perceptron (MLP) to discriminate between 31 classes of leaves. The features have been applied individually as well as in combination to investigate how recognition accuracies can be improved. Experimental results demonstrate that the proposed approach is effective in recognizing leaves with varying texture, shape, size and orientations to an acceptable degree. Methodology for plant leaf recognition using shape and texture features is proposed.Features are made invariant to scaling and orientation of leaf images.Classification is done using two different types of neural classifiers.System is tested using both known and unknown classes of leaf images.System is also designed to handle images with small amounts of deformations.

Adaptive Firefly Algorithm for nonholonomic motion planning of car-like system

This paper provides a novel approach to design an Adaptive Firefly Algorithm using self-adaptation of the algorithm control parameter values by learning from their previous experiences in generating quality solutions. Computer simulations undertaken on a well-known set of 25 benchmark functions reveals that incorporation of Q-learning in Firefly Algorithm makes the corresponding algorithm more efficient in both runtime and accuracy. The performance of the proposed adaptive firefly algorithm has been studied on an automatic motion planing problem of nonholonomic car-like system. Experimental results obtained indicate that the proposed algorithm based parking scheme outperforms classical Firefly Algorithm and Particle Swarm Optimization with respect to two standard metrics defined in the literature.

Query Expansion Using Term Distribution and Term Association

Good term selection is an important issue for an automatic query expansion (AQE) technique. AQE techniques that select expansion terms from the target corpus usually do so in one of two ways. Distribution based term selection compares the distribution of a term in the (pseudo) relevant documents with that in the whole corpus / random distribution. Two well-known distribution-based methods are based on Kullback-Leibler Divergence (KLD) and Bose-Einstein statistics (Bo1). Association based term selection, on the other hand, uses information about how a candidate term co-occurs with the original query terms. Local Context Analysis (LCA) and Relevance-based Language Model (RM3) are examples of association-based methods. Our goal in this study is to investigate how these two classes of methods may be combined to improve retrieval effectiveness. We propose the following combination-based approach. Candidate expansion terms are first obtained using a distribution based method. This set is then refined based on the strength of the association of terms with the original query terms. We test our methods on 11 TREC collections. The proposed combinations generally yield better results than each individual method, as well as other state-of-the-art AQE approaches. En route to our primary goal, we also propose some modifications to LCA and Bo1 which lead to improved performance.

A Modified History Based Weighted Average Voting with Soft-Dynamic Threshold

Voting is a widely used fault-masking technique for real time systems. Several voting algorithms exist in literature. In this paper, a survey on the few existing voting algorithms is presented and a modified history based weighted average voting algorithm with soft-dynamic threshold value is proposed with two different weight assignment techniques, which combines all the advantages of the surveyed voting algorithms but overcomes their deficiencies. The proposed algorithm with both type of weight assignment techniques, gives better performance compared to the existing history based weighted average voting algorithms in the presence of intermittent errors. In the presence of permanent errors, when all the modules are fault prone, the proposed algorithm with first type of weight assignment technique gives higher availability than all the surveyed voting algorithms. If at least one module is fault free, this algorithm gives almost 100 % safety and also higher range of availability than the other surveyed voting algorithms.

Improving Pseudo Relevance Feedback in the Divergence from Randomness Model

In an earlier analysis of Pseudo Relevance Feedback (PRF) models by Clinchant and Gaussier (2013), five desirable properties that PRF models should satisfy were formalised. Also, modifications to two PRF models were proposed in order to improve compliance with the desirable properties. These resulted in improved retrieval effectiveness. In this study, we introduce a sixth property that we believe PRF models should satisfy. We also extend the earlier exercise to Bo1, a standard PRF model. Experimental results on the robust, wt10g and gov2 datasets show that the proposed modifications yield improvements in effectiveness.

Plant Leaf Recognition Using Ridge Filter and Curvelet Transform with Neuro-Fuzzy Classifier

The current work proposes an innovative methodology for the recognition of plant species by using a combination of shape and texture features from leaf images. The leaf shape is modeled using Curvelet Coefficients and Invariant Moments while texture is modeled using a Ridge Filter and some statistical measures derived from the filtered image. As the features are sensitive to geometric orientations of the leaf image, a pre processing step is performed to make features invariant to geometric trans-formations. To classify images to pre-defined classes, a Neuro fuzzy classifier is used. Experimental results show that the method achieves acceptable recognition rates for images varying in texture, shape and orientation.

Recognition of whole and deformed plant leaves using statistical shape features and neuro-fuzzy classifier

This paper proposes a methodology for recognition of plant species by using a set of statistical features obtained from digital leaf images. As the features are sensitive to geometric transformations of the leaf image, a pre processing step is initially performed to make the features invariant to transformations like translation, rotation and scaling. Images are classified to 32 pre-defined classes using a Neuro fuzzy classifier. Comparisons are also done with Neural Network and k-Nearest Neighbor classifiers. Recognizing the fact that leaves are fragile and prone to deformations due to various environmental and biological factors, the basic technique is subsequently extended to address recognition of leaves with small deformations. Experimentations using 640 leaf images varying in shape, size, orientations and deformations demonstrate that the technique produces acceptable recognition rates.

A comparison of several nonlinear filters for ballistic missile tracking on re-entry

In this paper, ballistic missile tracking on re-entry has been considered. Position, velocity and ballistic coefficient are considered as states of the target. As no prior information about the target shape, mass and area is available, ballistic coefficient of the target has been considered as a state variable. Different type of nonlinear filters such as ensemble Kalman filter (EnKF), unscented Kalman filter (UKF), Gauss-Hermite filter (GHF), sparse-grid Gauss-Hermite filter (SGHF), cubature quadrature Kalman filter (CQKF) are used to estimate states of the target. The performance of all the above mentioned filters has been compared on the basis of estimation accuracy, computational time and missed distance. The 3 points GHF provides lowest miss distance at reasonably low computational time; so GHF-3 is recommended for this problem.

Fractional order sliding mode control based chaos synchronization and secure communication

The present work demonstrates a novel fractional order sliding mode controller for synchronization of fractional order chaotic systems and its use in secure communication. First, the existence of the chaotic behaviour is ensured for the autonomous chaotic system under consideration and then a sliding mode controller is designed in order to make the system follow a designated state trajectory. This idea is then extended towards the making of a chaotic transmitter and a chaotic receiver. It is shown that, when synchronized, this can be used to securely transmit an analogue signal.

Tracking of ballistic target on re-entry using ensemble Kalman filter

In this work, ground radar based ballistic target tracking problem in endo-atmospheric re-entry phase with unknown ballistic coefficient has been solved using ensemble Kalman filter (EnKF). EnKF, a powerful tool in nonlinear estimation, is being extensively used by meteorologist but almost unknown to target tracking community. Performance improvement, and computational burden of EnKF with increasing ensemble size have been studied. Performance of EnKF has been compared with most popular extended Kalman Filter (EKF) in terms of biasness, estimation accuracy, and computational efficiency. The simulation results reveal that the estimation accuracy of EnKF with sufficient ensemble size is much better than EKF.