THE METHOD OF FINDING THE SPAM IMAGES BASED ON THE HASH OF THE KEY POINTS OF THE IMAGE

Abstract

The article deals with the problems of the existing image recognition algorithms and also presents a method allowing to minimize the drawbacks of existing algorithms, and even to surpass them in certain moments. The proposed method is based on the description of the key points of the image by using perceptual hash. As a demonstration of application of the developed algorithm it is proposed to make the spam filter for images. The purpose of the spam filter is to separate, to divide the images from the collection into certain classes. The core of the spam filter is the developed method. In its conclusion the article contains information that shows the accuracy and performance of the proposed method in comparison with existing methods.