In this paper, we propose Unimodal Biometric Identification System Based on Ensemble machine learning techniques such as Stacking, Boosting and Bagging. The article framework gives the comparative analysis between Ensemble techniques adopting different classifiers (Geometric based classifiers-KNN, SVM, NN and Decision tree basedRF, RPART) and the best feature extraction algorithms such as Texture based-LPQ, Appearance basedICA1. The experimentation is carried out on physiological biometric traits adopting the standard benchmark databases such as AR facial database and Poly-U fingerprint database. The objective of the paper is to understand whether concentrating on discriminating feature extraction algorithms or performing extensive computation on ensemble techniques with different classification models would contribute greatly in performance of the system.

One of the major difficulties encountered by current face recognition systems deception in the problems of handling partial face recognition such as varying poses, illumination, light scattering, diffused faces i. e., recognition of faces in random in-depth rotations. The face image differences caused by rotations are often superior to the inter-person differences used in distinctive identities. Initially we have taken full face as training set, part of the images are make it as rotational, flip horizontal, flip vertical, scaled up, scaled down facial images which are considered as testing set (for both we use AR database) then convert the colored images in to gray level images. Secondly normalize the histogram of an image it represents the relative frequency of occurrence of the various gray levels in the image. Subsequently, Testing cases are divided in to four ratio cases like as 70: 30, 60: 40, 50: 50,80: 20 using this bifurcation cases of images we will get different recognition results. Simultaneously we have computed histogram of the normalized images and we got features in the form of matrix or vector representation and then compare the values of different testing cases. Finally, matching the features using template matching. The proposed algorithm has been testing with AR dataset and results obtained are satisfactory.

In this paper, we are proposing new ensemble strategy for classification of lung nodules based on their malignancy ratings. The procedure we followed is simpler. In the first step, we construct different homogenous ensemble models such as bagged decision tree (BaDT), boosted decision tree (BoBT), and random subspace–based decision tree (RSSDT). In the next step, we combine previously constructed models with voting scheme to yield ensemble of homogenous ensemble of classifiers. We also examine the behavior of our method for heterogeneity in the system. This is done by constructing ensemble of heterogeneous ensemble of classifiers. For this, we have also considered bagged KNN (BaKNN), boosted KNN (BoKNN), bagged PART (BaPART), and boosted PART classifier (BoPART). The results we are obtaining from our strategy are significant compared to homogenous ensemble model.

In this work, we are proposing a new machine learning strategy for classification task for imbalanced data. We are using lung image data by Lung Image Database Consortium (LIDC), since LIDC data is a better example for imbalanced dataset. In this work we are using sufficiently large dataset which contains 4,532 nodules extracted from CT images. Later we consider 55 low level nodule image features and radiologists ratings for experiments. This work is being dealt in two stages. (1) data level learning and (2) algorithm level learning. In first stage, we are balancing the dataset prior to classification process. We are using resampling approach for this task. In second stage, we are using ensemble of classifiers to predict lung nodule rating. We are using wide range of classifier models for constructing an ensemble. We use Bagged Decision Tree, naïve Bayes, Boosted Decision Trees, and Support Vector Machine (SVM) in a classifier library. Stacking algorithm is used to combine the different classifier models in library to construct higher level ensemble. We are evaluating the performance of our model on five metrics: Accuracy, precision, recall, F-score and Kappa statistics. Results show that our method yields much improved scores as we are refining at both, data level and algorithm level.

The acquired image will has different levels of noise pollution and image distortion in Handwritten Chinese character recognition document image processing. For these situations the accurate and fast edge detection method is an important prerequisite for the recognition results. The widely used edge detection algorithms such as: first derivative-based edge detection method, second derivative edge detection method, canny operator, mathematical morphology edge detection and fuzzy edge detection method. In this paper according to our database of handwritten Chinese characters a lot of experiments to compare the excellent of various algorithms in the field of Chinese character handwriting recognition application. These algorithms are sensitive to image noise and we nowadays discuss new edge detection for the theory of wavelet edge detection method and genetic algorithms. According to our experimental results we explore the new edge detection method to improve the edge detection, image segmentation accuracy and quickness.

Text block verification is important in enhancing the text detection accuracy in natural scene images because it is hard to develop general or objective heuristics to differentiate text and non-text block. In this paper, we propose new objective heuristics to verify the blocks detected by the text detection method based on angle information and proximity of the blocks. The angle for the detected block is computed using PCA to find the direction of the text block. In the same way, the proximity between pixels in the detected block is estimated to find closeness between pixels. Then the method combines these two heuristics to verify the text block to obtain a better result. We conduct experimental results on different databases to show that the performance of the text detection method increases in terms of recall, precision and f-measure with the text block verification methods. The database includes benchmark database ICDAR-2003 competition data, our own data captured by high resolution camera and captured by low resolution mobile camera.

An electrocardiogram (ECG) is a bioelectrical signal which records the heart’s electrical activity versus time. The interpretation of ECG signal is an application of pattern recognition. The techniques used in this paper comprise: signal preprocessing, R peak detection, QRS reconstruction, RR interval detection, feature extraction and linear classifier model versus ensemble of classifier model. The processed signal source came from the Massachusetts Institute of Technology Beth Israel Hospital (MIT-BIH) arrhythmia database which was developed for research in cardiac electro-physiology. The results of recognition rates are compared to find a better structure for ECG classification. Among different classifier model, it was found that ensemble of classifier with DECORATE meta-learner model possessed the best performance with highest recognition rate of 90.36% for cardiac conditions and moderate level of agreement between computerized prediction and cardiologist interpretation. Based on this result, the method of using important ECG features plus a suitable ensemble of classifier model outperforms the single classifier model and which can increase the testing speed and the accuracy rate.

Reverse Engineering is a process of analyzing the subject system to identify its components, and relationships so as to represent the system at higher levels of abstractions to help developers understand the system for later maintenance and enhancement. Many software systems deployed in both military and industrial domains are very complex and comprise of multiple threads of control. Such systems are expensive and time consuming to build and must be evolved to meet new challenges. Hence, the challenge lies in discovering the information about the threads and their interactions for later maintenance and enhancement activities. In this paper, we have proposed a static model that analyses C++ multithreaded LINUX source code, extracts information about threads, their interactions and record their understanding. The extracted information is used for pictorial presentation and program comprehension. The details obtained from multithreaded source code is helpful in clear understanding of architecture of threads and their interactions resulting into reduced effort in maintenance and enhancement of software.

Thispaperdescribes thecharacter recognition process from printed documents containing KannadaandEnglish text. Kannada isthefifth mostpopular language inIndia andEnglish isthemostpopular language intheworld. Kannadaisthe language spoken bymorethan60million people ofSouth India andEnglish isthesecondofficial language atvarious government organizations through outIndia. Theproposed character recognizer isbasedonthefisher linear discriminant (FLD)analysis. Itisusually performed toinvestigate differences amongmultivariate classes, todetermine whichattributes discriminate theclasses, andtodetermine themostparsimonious waytodistinguish amongclasses. Theproposed system istested on various fonts, degraded characters, noisycharacters of KannadaandEnglish. Theoverall accuracy oftheproposed system is96.1%.