Search

Abstract: In this paper, we identify some of the limitations of current-day shape-matching techniques. We provide examples of how contour-based shape-matching techniques cannot provide a good match for certain visually similar shapes. To overcome this limitation, we propose a perceptually motivated variant of the well-known shape context descriptor. We identify that the interior properties of the shape play an important role in object recognition and develop a descriptor that captures these interior properties. We show that our method can easily be augmented with any other shape-matching algorithm. We also show from our experiments that the use of our descriptor can significantly improve the retrieval rates. [Copyright &y& Elsevier]

Abstract: Let denote the symmetric group on and the stabilizer subgroup of . We derive algorithms for computing Fourier transforms of left and right -invariant signals that require a total of additions and scalar multiplications. Furthermore we show that the convolution of such signals can also be computed in time linear in . [Copyright &y& Elsevier]

Describes a spatially adaptive demosaicing algorithm that is based on the Jacobian matrix of the color map and neighborhood voting. Definition of demosaicing; Computational requirements of the algorithm; Improvement of the algorithm on published algorithms in terms of complexity or image quality.

Digital camera sensors usually contain defective pixels that either arise during the manufacturing process, or develop over time. These defects though small in number, are very noticeable to an experienced eye when looking at images captured from the camera. As a result it becomes extremely important for the sensor manufacturing companies to develop methods which would reduce these manufacturing errors. There have been a number of algorithms proposed for online detection of sensor defects in the literature, but there is no standard method of fairly comparing algorithms given that each algorithm uses thresholds in a different way. This paper presents a novel idea which makes use of the Receiver Operating Characteristics (ROC) curve for comparison among the various bad pixel correction algorithms. We demonstrate how the characteristics of the ROC curve can be used for determining the effectiveness of the algorithms and hence help us to decide which of the algorithms perform better than the others. Experimental results comparing two such bad pixel detection algorithms are provided1. [ABSTRACT FROM AUTHOR]