A study on image hashing techniques for implementing deduplication

Abstract

Image deduplication is an approach to avoid duplicate images to be stored in cloud, which is currently required to increase the efficient utilization of cloud storage. Most of the times, the images uploaded by different users of cloud storage service need not to be unique. There is a possibility of storing same images by different users. This may lead to poor usage of cloud storage. To find duplicates, various image hashing techniques are widely used. The purpose of this paper is to review various image hashing methods which are used to generate hash value of the images. Using the hash value, we can conclude the uniqueness of the image. This can be achieved by using the “collision resistance” property of hash value. To measure the feasibility, the general necessity for image deduplication in cloud is discussed in this work. Additionally, the performance of few selected image hashing methods is compared with respect to robustness and discrimination.