Back to results
US 8,805,104 B1Granted
Selecting example-based predictors based on spatial continuity
AssigneeGoogle Inc.
InventorsMichele Covell, Vivek Kwatra, Mei Han, Saurabh Mathur, Shumeet Baluja
Life of this patent
Estimating the term…
Abstract
An image processing system is provided for encoding images based on example-based compression. The system selects a set of candidate dictionary predictor entries to encode a portion of an image based at least in part on the neighbors of the portion. The spatial continuity between portions of the image is exploited by the image processing system by selecting corresponding dictionary predictor entries that have the same offset vector as the portion of the image and its neighboring portions.
Claims20
A method, comprising:
- identifying, by a device comprising a processor, an image to encode, the image comprising a plurality of portions that collectively represent the image;
- selecting, from dictionary predictor entries, one or more dictionary predictor entries to encode a portion of the image, the selection being based at least in part on spatial locality of the portion of the image relative to one or more neighboring portions of the image, the one or more neighboring portions being directly adjacent to the portion of the image;
- encoding the portion of the image using the one or more dictionary predictor entries to generate encoded information;
- evaluating quality of the encoding using the one or more dictionary predictor entries;
- and
- selecting one of the one or more dictionary predictor entries based on the quality of the encoding,
- wherein the selecting the one or more dictionary predictor entries to encode the portion of the image comprises:
- identifying, for a randomly selected dictionary predictor entry, a spatial dictionary predictor entry associated with a reference image that corresponds to the randomly selected dictionary predictor entry, the spatial dictionary predictor entry being positioned within the reference image at an offset vector from a position of the randomly selected dictionary predictor entry within the reference image, and the offset vector being equivalent to a vector between the portion of the image and at least one of the one or more neighboring portions.
5 more elements
A non-transitory computer-readable storage medium storing executable code that, in response to execution, cause a device to perform operations comprising:
- identifying an image to encode, the image comprising a plurality of portions that collectively represent the image;
- selecting, from dictionary predictor entries, one or more dictionary predictor entries to encode a portion of the image, the selection being based at least in part on spatial locality of the portion of the image relative to one or more neighboring portions of the image, the one or more neighboring portions being directly adjacent to the portion of the image;
- encoding the portion of the image using the one or more dictionary predictor entries to generate encoded information;
- evaluating quality of the encoding using the one or more dictionary predictor entries;
- and
- selecting one of the one or more dictionary predictor entries based on the quality of the encoding,
- wherein the selecting the one or more dictionary predictor entries to encode the portion of the image comprises:
- identifying, for a randomly selected dictionary predictor entry, a spatial dictionary predictor entry associated with a reference image that corresponds to the randomly selected dictionary predictor entry, the spatial dictionary predictor entry being positioned within the reference image at an offset vector from a position of the randomly selected dictionary predictor entry within the reference image, and the offset vector being equivalent to a vector between the portion of the image and at least one of the one or more neighboring portions.
5 more elements
An apparatus, comprising:
- a processor;
- and
- a non-transitory computer-readable storage medium storing executable code that, in response to execution by the processor, cause the apparatus to perform operations comprising:
- identifying an image to encode, the image comprising a plurality of portions that collectively represent the image;
- selecting, from dictionary predictor entries, one or more dictionary predictor entries to encode a portion of the image, the selection being based at least in part on spatial locality of the portion of the image relative to one or more neighboring portions of the image, the one or more neighboring portions being directly adjacent to the portion of the image;
- encoding the portion of the image using the one or more dictionary predictor entries to generate encoded information;
- evaluating quality of the encoding using the one or more dictionary predictor entries;
- and
- selecting one of the one or more dictionary predictor entries based on the quality of the encoding,
- wherein the selecting the one or more dictionary predictor entries to encode the portion of the image comprises:
- identifying, for a randomly selected dictionary predictor entry, a spatial dictionary predictor entry associated with a reference image that corresponds to the randomly selected dictionary predictor entry, the spatial dictionary predictor entry being positioned within the reference image at an offset vector from a position of the randomly selected dictionary predictor entry within the reference image, and the offset vector being equivalent to a vector between the portion of the image and at least one of the one or more neighboring portions.
8 more elements
Dependencies are inferred from the claim text.