Methods of representing and analysing images
A method of representing and analysing images comprises producing a plurality of descriptors of an image at one or more scales and for one or more color channels, said descriptors capturing color content and interrelation information within the regions, and associating the descriptors in a plurality of ways based on their characteristics such as scale, color channel, feature semantics, and region, and comparing such representations of images to assess the similarity of images.
1. A method of representing at least one image, using one or more processors, comprising:
deriving, using one or more processors, a plurality of first sub-descriptors representing colour information for a plurality of first windows of said image and a plurality of second sub-descriptors representing colour interrelation information for a plurality of second windows of said image, wherein said first windows comprise windows defined in different sizes.
2. The method of claim 1 further comprising
producing a reduced image of size 32×32 pixels corresponds to said image,
wherein said reduced image is used for said process of deriving the first and second sub-descriptors.
3. The method of claim 1 wherein said second windows comprise windows overlapping each other and windows which do not overlap each other.
4. The method of claim 1 wherein different neighbourhood parameters or sub-descriptor extraction processes are employed at different locations in said image.
5. The method of claim 1 wherein different processes for deriving sub-descriptors are used for different regions.
6. The method of claim 1 wherein at least one sub-descriptor element for a region, or each or all of a plurality of sub-descriptor elements for a region, is or are derived using only a subset of pixels in said region.
7. The method of claim 1 further comprising associating a plurality of second sub-descriptors or sub-descriptor elements are associated on the basis of one or more of colour interrelation.
8. The method of claim 1 wherein all the descriptor elements are derived for a single colour channel.
9. The method of claim 1 using colour space YCrCb.
10. The method of claim 1 comprising deriving descriptor elements for luminance channel Y.
11. The method of claim 1 wherein colour is derived from averages of pixel values of said first windows and colour interrelation is derived from differences based on pixel values of said second windows.
12. A method of representing a group of images using one or more processors, comprising
deriving a representation of each image using a method of claim 1 ; and
using said representations to represent said group of images, comprising deriving, using one or more processors, a function of said representations to represent the group of images, wherein said function includes a statistical measure.
13. A method of comparing images or groups of images using one or more processors comprising
comparing, using one or more processors, respective representations derived using the method of claim 1 , wherein the comparison involves comparing descriptors which correspond according to characteristics.
14. A method of comparing images or groups of images using one or more processors comprising
comparing, using one or more processors, respective representations derived using a method of claim 1 , wherein the comparison involves comparing descriptors which correspond according to characteristics.
15. An apparatus of representing at least one image comprising:
a memory; and
one or more processors configured to perform processes including deriving a plurality of first sub-descriptors representing colour information for a plurality of first windows of said image and a plurality of second sub-descriptors representing colour interrelation information for a plurality of a second windows of said image,
wherein said first windows comprise windows defined in different sizes.
16. An apparatus of representing a group of images comprising:
a memory; and
one or more processors configured to perform processes including
deriving a representation of each image using an apparatus of claim 15 , and using said representations to represent said group of images, comprising deriving a function of said representations to represent the group of images for example, wherein said function comprises a statistical measure such as average, median, mean, covariance or the like.
17. An apparatus of comparing images and/or groups of images comprising:
a descriptor database storing descriptors of images or groups of images; and
one or more processors configured to perform processes including
comparing respective representations derived using an apparatus of claim 15 ,
wherein the comparison involves comparing descriptors which correspond according to characteristics.
18. An apparatus of comparing images and/or groups of images comprising:
a descriptor database storing descriptors of images or groups of images; and
one or more processors configured to perform processes including
comparing respective representations derived using an apparatus of claim 16 ,
wherein the comparison involves comparing descriptors which correspond according to characteristics.