Back to results
US 2005/0031218 A1Published application
Tile boundary artifact removal for arbitrary wavelet filters
AssigneeRicoh Company, Ltd.
InventorsKathrin Berkner, Edward Schwartz
Life of this patent
Estimating the term…
Abstract
A method for processing data is described. In one embodiment, the method includes decompressing a plurality of sets of compressed data, including performing detiling on a first set of compressed data using neighbor information from at least one other set of compressed data, and recombining the plurality of decompressed tiles into an image.
Claims22
1-43. (Canceled)
A method for processing data, the method comprising:
- receiving high and low pass wavelet coefficients and neighboring data;
- and
- modifying coefficients in at least one of the highpass and lowpass wavelet coefficients at a decomposition level using neighboring highpass and lowpass wavelet coefficients at the decomposition level.
A wavelet-based system, comprising:
- an input to receive high and low pass wavelet coefficients and neighboring data;
- and
- one or more decompressors to modify coefficients in at least one of the highpass and lowpass wavelet coefficients at a decomposition level using neighboring highpass and lowpass wavelet coefficients at the decomposition level.
An apparatus for processing data, comprising:
- means for receiving high and low pass wavelet coefficients and neighboring data;
- and
- means for modifying coefficients in at least one of the highpass and lowpass wavelet coefficients at a decomposition level using neighboring highpass and lowpass wavelet coefficients at the decomposition level.
A method, comprising:
- compute scaling coefficients for a tile wavelet decomposition at level J-1 by applying a tile inverse transform to scaling and wavelet coefficients at level J;
- compute an approximation of phase coefficients at level J by applying a full-frame forward transform to a shifted sequence of s coefficients at level J-1;
- computing an approximation of the phase coefficients at level J-1 by applying an inverse full-frame wavelet transform to the shifted sequence of s coefficients;
- computing an approximation of the s coefficients at level J by applying a forward tile transform to the approximation of the phase coefficients.
1 more element
A method, comprising:
- decoding coded data to generate a plurality of coefficients;
- and
- performing a multi-level inverse wavelet transform on the plurality of coefficients to generate modified coefficients, wherein performing the multi-level inverse wavelet transform comprises generating approximations for coefficients near tile boundaries, wherein an approximation of full-frame scaling coefficients sJ is {overscore (s)}J.
A method, comprising:
- receiving wavelet coefficients;
- and
- modifying coefficients of the wavelet coefficients at a decomposition level using coefficient phase of coefficients of neighboring tiles.
A method, comprising:
- correcting tile highpass coefficients at a scale using neighboring lowpass coefficients at the scale based on a target image obtained by
- generating, at each reconstruction level J, using a full-frame wavelet decomposition of an original image modified only at highpass coefficients at level J that are affected by boundary corrections projected onto a scaling function associated with a low pass filter at locations affected by boundary conditions in a tile transform with modified coefficients set to zero.
A method comprising:
- performing detiling using coefficient phase of coefficients of a plurality of neighboring tiles
- recombining tiles to create a reconstructed image.
A method comprising:
- receiving wavelet coeffients;
- and
- performing a projection onto a scaling function associated with a lowpass wavelet filter at locations affected by boundary conditions in a tile transform.
Dependencies are inferred from the claim text.