Content adaptive resizer
AssigneeAlgolith Inc.
InventorsChon Le Dinh, Ngoc Nguyen
Life of this patent
Estimating the term…
Abstract
The present description discloses a method for resizing an input image having input pixels. The method comprises segmenting the input made into region types by assigning one of the region types to each one of the input pixels; and interpolating the input image according to the region type assigned to each one of the input pixels to produce a resized image having interpolated pixels.
Claims39
A method for resizing an input image having input pixels, said method comprising:
- segmenting said input image into region types by assigning one of said region types to each one of said input pixels:
- and
- interpolating said input image according to the region type assigned to each one of said input pixels to produce a resized image having interpolated pixels.
The method of claim 6, wherein a resizing ratio is defined by the size of the input image relative to the size of the resized image, and further wherein said interpolating further comprises using said resizing ratio for subdividing said FIR low-pass filter coefficients into polyphase filter coefficients used for said polyphase filtering.
An apparatus for resizing an input image having input pixels, said apparatus comprising:
- a segmentation unit for segmenting said input image into region types by assigning one of said region types to each one of said input pixels;
- and
- an interpolation unit comprising a filter for filtering said input pixels according to the region type assigned to each one of said input pixels to produce a resized image having interpolated pixels.
A method for segmenting into region types an input image having input pixels, said method comprising:
- Identifying said region types in said input image using said input pixels, and
- assigning one of said region types to each one of said input pixels to thereby segment said image into said region types.
An apparatus for segmenting into region types an input image having input pixels, said apparatus comprising:
- a plurality of directional gradient operators for at least one of a) determining a frequency value for input pixels in at least one of said region types, and b) determining a direction value for input pixels in at least one of said regions, and for identifying said region types in said input image using said input pixels;
- and
- an assignment unit using at least one of said frequency value and said direction value for assigning one of said region types to each one of said input pixels to thereby segment said image into said region types.
Dependencies are inferred from the claim text.