Abstract: Interval Bezier curve are new representation forms of parametric curves. Using this new representation, the problem of lack of robustness in all state-of-the art CAD systems can be largely overcome. In this paper this concept has been discussed to form a new curve over rectangular domain such that its parameter varies in an arbitrary range [a,b] instead of standard parameter [0,1]. Where a and b are real and, we also want that curve gets generated within the given error tolerance limit. The four fixed Kharitonov's polynomials (four fixed Bezier curves) associated with the original interval Bezier curve are obtained. A new parameterization is applied to the four fixed Kharitonov's polynomials (four fixed Bezier curves). Finally, the required interval control points are obtained from the fixed control points of the four fixed Kharitonov's polynomials. Subdividing a parametric interval Bezier curve into two interval segments is also presented. The two interval segments have the same shape as the original interval Bezier curve, but they are defined by more entities (interval control points or interval vectors) thereby making it possible to fine-tune the interval Bezier curve. Using matrix representation, it has been shown how to determine the control polygon that covers an arbitrary interval [a, b] of the original interval Bezier curve. Numerical examples are included in order to demonstrate the effectiveness of the proposed method.

Abstract: Malaria, a life threatening disease named in 1740, is largely a geographical disease, endemic to tropical climates. It is caused by the protozoans called plasmodium. The infected female Anopheles mosquitoes, also known as malaria vectors spread the parasites to people through their bites. Malaria diagnosis involves identifying malaria parasites in patient blood. Contemporary malaria diagnosis techniques basically depend on microscopic analysis of Giemsa-smeared thin and thick films of blood. However due to inherent technical limitations and the number of steps required in manual assessment, this diagnostic method is time consuming and prone to human error. This situation has prompted an increasing interest in finding technological solutions to carrying out the diagnosis automatically. This paper presents an efficient approach for automatic malaria detection with fuzzy based color segmentation, fractal feature extraction and ART neural network classification. The process starts with converting the input image to gray scale, LAB and HSV. L and B planes from LAB image and S plane from HSV image are extracted for identification of parasites. Fuzzy based segmentation technique is followed. Further color features and fractal features are extracted. Based on these features, malaria parasites in an image can be identified. Identification is done using four classifiers- Adaptive Resonance Theory (ART) based neural network, Support vector machine (SVM), Neural Network based Back propagation Feed Forward (NN-BPFF) and k-Nearest Neighbor (k-NN). These classifiers automatically classify the images as malaria and non-malaria. A Performance Evaluation toolbox has been designed and developed for the malaria parasite classification and comparative analysis has been done with all the four classifiers. Best performance of 98.52% has been recorded for ART classifier Receiver Operating Characteristic Curve (ROC).

Abstract: This paper presents a description of a novel mining system which mines the different occurrences (instances) of the same object from a video file. The framework of the system consists of four steps: segmenting the video file into stable tracks, extracting objects and their features from the tracks, grouping these tracks into clusters based on their residing objects, and finally mining the instances of each object in the shared pool of configurable computing resources within cloud environment for more security. The paper also presents a critique and feedback for the system and proposes an idea to improve its performance.