Geometry implicit sampler for polynomial surfaces over freeform

Details Geometry implicit sampler for polynomial surfaces over freeform Geometry implicit sampler for polynomial surfaces over freeform

1997-07-07

1999-06-29

Shalwala, Bipin

Image analysis

Image compression or coding

Adaptive coding

382243, 382291, 382151, G06K 936, G06K 946, G06K 900

Patent

active

059179538

ABSTRACT:
In two-dimensional model based image coding, an image is partitioned into domains, and the pixels of each domain are modeled with a polynomial intensity function. An image coded in this fashion is comprised of a description of the domain boundaries and individual descriptions of polynomial intensity functions. In order to trade rate for distortion it is desirable that the accuracy of the descriptions be proportional to their length. A polynomial description that is natural for trading rate for distortion is a sequence of samples of its generated surface. If the locations of the samples are implicitly determined from domain shape, an n term polynomial description is comprised of n sample values, and rate is traded for distortion via sample value quantization. In order to bound the sample values to the dynamic range of the image, their locations must be constrained to lie within their associated domain. The present invention is a method for implicitly determining appropriate locations within a freeform two-dimensional domain at which to sample a polynomial surface over the domain such that when the surface is recovered from the samples it accurately reflects the sampled surface and is minimally sensitive to sample value changes. The method determines sample locations solely from domain shape and the expected number of polynomial terms. The number of sample locations determined is nominally equivalent to expected number but is reduced if appropriate locations cannot be found. The method is comprised of two orthogonal one-dimensional samplers and a reduced two-dimensional sampler largely confined to the associated domain's periphery. Algorithmic complexity is O(.sqroot.N) where N is the number of pixels in the sampled domain.

REFERENCES:
patent: 5734756 (1998-03-01), Sherman et al.

Affiliated with

Also associated with

Rating

Geometry implicit sampler for polynomial surfaces over freeform does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Geometry implicit sampler for polynomial surfaces over freeform , we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Geometry implicit sampler for polynomial surfaces over freeform will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-1383798