Image descriptor for image recognition

Abstract

A method of deriving a representation of an image by processing signals corresponding to the image, the method comprises deriving a function of the image, where the function of a translated, scaled or rotated version of the image is a translated or scaled version of the function of the image, and using a plurality of frequency components of a frequency representation of the function to derive a representation of the image.

Description

technical field [0001] The present invention relates to methods and devices for representing images, and methods and devices for comparing or matching images (eg for searching or verification, etc.). Background technique [0002] Extensive image databases exist, and it is often necessary to find copies of a particular image or copies of modified versions of that image. Such functionality may be required, for example, in rights management or for search purposes. The same image on different websites can represent similar content and thus provide a valuable search lead. Another important application is the identification and association between "film" photographs and negatives and other digital archives. [0003] A hashing algorithm such as MD5 can be used to detect identical images. A hash function generates a value based on content; if two images are identical, they will have the same hash value. However, traditional methods such as MD5 are bit-sensitive, meaning that a c...

AI Technical Summary

Problems solved by technology

[0024] ●The false acceptance rate is too high for effective use with large databases

Images