IMAGE PROCESSING, PATTERN RECOGNITION

Efficient linear local features based copy-move detection algorithm

A. V. Kuznetsov^a, V. V. Myasnikov<sup>ab

a</sup> S.P. Korolyov Samara State Aerospace University (National Reseach University)
^b Image Processing Systems Institute of the RAS

Abstract: In this paper we propose a new algorithm to detect copy-move matching parts on digital images. The algorithm uses a representation of the image’s analyzed region as the value of a specially selected hash function. Actually, the hash function is constructed using the effective linear local features. The developed algorithm is compared with existing algorithm based on CRT, the recommendations are given.

Keywords: digital image, duplicate, hash function, hash table, histogram, prime Galois field linear local features.

Received: 11.09.2013