Title :
Mean curvature evolution and surface area scaling in image filtering
fDate :
5/1/1997 12:00:00 AM
Abstract :
Representing the image as a surface, an inhomogeneous diffusion algorithm is developed, evolving the surface at a speed proportional to its mean curvature, reducing noise while preserving image structure. An adaptive scaling parameter increases the speed of the diffusion. The properties of a discrete algorithm are demonstrated experimentally
Keywords :
adaptive filters; diffusion; edge detection; filtering theory; image enhancement; image representation; nonlinear filters; adaptive nonlinear scaling parameter; discrete algorithm; image filtering; image representation; inhomogeneous diffusion algorithm; mean curvature evolution; noise reduction; surface area scaling; Anisotropic magnetoresistance; Differential equations; Filtering algorithms; Image processing; Lead isotopes; Level set; Noise reduction; Proportional control; Space heating;
Journal_Title :
Image Processing, IEEE Transactions on
