DCT and perceptual hashing based to identify texture anti-counterfeiting tag
Yuji Li · Jisuanji yingyong yanjiu · 2014
In view of the problem that texture anti-counterfeiting tags were difficult to identify to the naked eye in the case of poor eyesight or insufficient light,this paper put forward an algorithm based on DCT,perceptual hashing and human visual characteristics for the identification of texture anti-counterfeiting tags. Firstly,the method extracted the visual feature vectors of texture image through DCT and perceptual hashing,and then created a database of texture images' feature vectors. At last the scheme compared the normalized cross-correlation( NC) between the feature vectors to realize automatic identification of texture tags. The results of experimental study demonstrate that the proposed algorithm has strong robustness against common and geometric attacks. In addition,compared with the existing texture anti-counterfeiting technology,this algorithm can reduce the database storage space occupied and has better practicability.