Jayant Shah

This paper is essentially another version of Horikawa's analysis [1]. The purpose here is to obtain a complete list of ("minimal") projective degen-erations analogous to the list obtained by Morrison [2] by the analytic method of Kulikov, Persson and Pinkham. (Morrison's list is reproduced ...

Page 1. SEGMENTATION BY NONLINEAR DIFFUSION Jayant Shah Mathematics Department, Northeastem University, Boston, Mass. 021 15 ABSTRACT' One of the basic problems in Computer Vision is the problem of segmenting an image into meaningful re-gions. ...

Abstract A new method for determining skeletons of 3D shapes is described. It is a combination of the approach based on the "grass-fire" technique and Zhu's approach based on first finding portions of the shape where its width is approximately constant. The method specifically does not re-quire presmoothing of the shape and is robust in the presence of noise. In

A new approach is presented for recovering shapes from noisy images. First, an edge-strength function, v, at every pixel is determined by implementing a segmenta- tion functional. Then, the zero-crossings of the laplacian of the smoothed image are allowed to evolve under the influence of v. Each point on the zero-crossings moves in the direction of the normal with a

A system of coupled differential equations that learns priors for modeling “preattentive” textures is formulated. Learning is driven by the feature residuals computed from the observed values and the values calculated by the system from a synthesized image that is ...

During the past decade, curve evolution has been applied to shape recovery, shape analysis, image smoothing, and image segmentation. Almost all of these applications are based on curve evolution which minimizes the total length of the curve. The curve moves ...

The method of curve evolution is a popular method for recovering shape boundaries. However isotropic metrics have always been used to induce the flow of the curve and potential steady states tend to be difficult to determine numerically, especially in noisy or ...

This paper presents a new approach to prior shape and appearance modeling for use in curve evolution-based seg-mentation. The new method is based on the unified use of feature distributions and allows the incorporation of coupled prior information about shape and appearance ...

Since lunar and solar parallax play a crucial role in predicting solar eclipses, the focus of this paper is on the computation of parallax. A brief history of parallax computation in India and China is traced. Predictions of solar eclipses based Nilakantha's Tantrasangraha are statistically analyzed. They turn
out to be remarkably accurate, but there is a pronounced bias towards predicting false positives rather than false negatives. The false positives occur more to the south of the ecliptic at northerly terrestrial latitudes and more to the north of the ecliptic at southerly latitudes. A very similar bias is found in Chinese astronomy providing another hint at possible links between Indian and Chinese astronomy.  The Chinese have traditionally used different values for the eclipse limit north and south of the ecliptic, perhaps to compensate for the southward bias.

This work provides a variational framework for fusing range and intensity data for recovering regularized surfaces. It is shown that this framework provides natural boundary conditions for the shape-from-shading problem, results in a new shape-from-shading formulation in the absence of range data, and provides a new fusion paradigm when range data is incorporated. The approach is demonstrated on simulated range and intensity images; error analysis with respect to the ground truth surface is presented. It is shown that the formulation performs well even in very noisy images