Method for automatic assessment of emphasis and appeal in consumer images

A method for determining the appeal of an image with respect to certain self-salient features, wherein said appeal is an assessment of the overall degree of importance, interest or attractiveness of a whole individual image, said method comprising the steps of:

• (a) obtaining a digital image corresponding to the image;
• (b) computing one or more quantities related to one or more self-salient features in the digital image, including one or more self-salient features pertaining to the content of the digital image;
• (c) processing said one or more quantities with a reasoning algorithm that is trained on the opinions of one or more human observers;
• (d) providing an output from the reasoning algorithm that assesses the appeal of the image, wherein said output comprises a score for assessing the overall degree of importance, interest or attractiveness of the whole individual image;
• and
• (e) ordering the image relative to other images in term of its score, so that an extreme score corresponds to an image having a greater appeal than the other images.

A method for determining the emphasis of an image with respect to certain features, wherein said emphasis is an assessment of the degree of importance, interest or attractiveness of an individual image relative to other images in a group of images, said method comprising the steps of:

• (a) obtaining a plurality of digital images corresponding to the group of images;
• (b) computing one or more quantities related to one or more features in each digital image, including one or more self-salient features pertaining to the content of each digital image;
• (c) processing the one or more quantities with a reasoning algorithm that is trained on the opinions of one or more human observers;
• (d) providing an output from the reasoning algorithm that assesses the emphasis of each image and computes an emphasis score for each image;
• and
• (e) ordering the images in the group of images in terms of their emphasis score, so that an extreme score corresponds to an image that is characterized as an emphasis image for the group.

A computer program product for assessing an image with respect to certain features, wherein said assessment is a determination of the overall degree of importance, interest or attractiveness of a whole image, said program product comprising:

• a computer readable storage medium having a computer program stored thereon for performing the steps of:
• (a) obtaining a digital image corresponding to the image;
• (b) computing one or more quantities related to one or more features in the digital image, including one or more features pertaining to the content of the digital image;
• (c) processing said one or more quantities with a reasoning algorithm that is trained on the opinions of one or more human observers;
• (d) providing an output from the reasoning algorithm that assesses the image, wherein said output comprises a score for assessing the overall degree of importance, interest or attractiveness of the whole image;
• and
• (e) ordering the image relative to other images in terms of its score, so that an extreme score corresponds to an image having a greater overall degree of importance, interest or attractiveness than the other image.

A method for determining the colorfulness of a digital image based on the chrominance component of its image pixels, said method comprising the steps of:

• (a) transforming the digital image to luminance/chrominance space;
• (b) dividing the chromaticity space into a plurality of bins;
• (c) associating each image pixel with one of the bins if its chrominance component lies within the bounds of that bin;
• (d) computing the level of saturation for each pixel in each bin;
• (e) determining the number of high saturation pixels in each bin;
• (f) identifying each bin where the number of high saturation pixels exceed a threshold as an active bin;
• and
• (g) determining the colorfulness from the number of active bins.

A method for determining compositional aspects of a digital image based on the outcome of main subject detection, said method comprising the steps of:

• (a) generating a main subject belief map indicative of a main subject in the digital image;
• (b) computing a variance measure of the pixels belonging to the main subject belief map;
• (c) determining a centrality measure of the main subject based on a distance between an image center and a centroid of a portion of the pixels belonging to the main subject belief map;
• (d) determining a compactness measure of the main subject by evaluating the proportion of a bounding box occupied by said portion of the pixels belonging to the main subject belief map;
• and
• (e) evaluating the goodness of the compositional aspects of the digital image by reference to the variance, centrality and compactness measures.