Image quality assessment
AssigneeXerox Corporation
InventorsLuca Marchesotti, Rodrigue Nkoutche
Life of this patent
Estimating the term…
Abstract
A computer-implemented system and method for predicting an image quality of an image are disclosed. For an input image, the method includes generating a first descriptor based on semantic content information for the image and generating a second descriptor based on aesthetic features extracted from the image. With a categorizer which has been trained to assign a quality value to an image based on first and second descriptors, a quality value is assigned to the image based on the first and second descriptors and output.
Claims23
A method for predicting an image quality of an image comprising:
- for an input image:
- generating a first descriptor based on semantic content information for the image;
- generating a second descriptor based on aesthetic features extracted from the image;
- with a categorizer which has been trained to assign a quality value to an image based on first and second descriptors, assigning a quality value to the image based on the first and second descriptors;
- and
- outputting the image quality value.
A system for generating an image quality assessment for an image comprising:
- memory which stores:
- a content feature extractor for generating a first descriptor based on semantic content information for an input image;
- an aesthetic feature extractor for generating a second descriptor based on aesthetic features extracted from the image;
- and
- a categorizer which has been trained to assign a quality value to an image based on the first and second descriptors;
- and
- a computer processor in communication with the memory for implementing the extractors and categorizer.
A method for predicting an image quality of an image comprising:
- receiving a set of training images, each image in the set having associated textual information describing the semantic content of the image, and an image quality value based on image quality information provided by a user;
- training a categorizer based on the labels of the images, their image quality values, and features extracted from the images;
- and
- for an input image having textual information but no image quality value:
- extracting features from image data of the image,
- with the categorizer, assigning a quality value to the image based on the textual information and extracted features, and
- outputting the image quality value.
Dependencies are inferred from the claim text.