A Content-based Image Retrieval using FFT & Cosine Similarity Coefficient

K. Singh, M. Ma, and D.-W. Park (Korea)


: FFT, Digital Signature, Content-Based Image Retrieval, Cosine Coefficient. .


Our mission was to get a highly efficient technique for Content Based Image Retrieval (CBIR). In this paper we address a new image retrieval technique that is based on digital signature of the image extracted from Fast Fourier Transformation (FFT). This technique is fully automatic, as it does not require any human intervention. Human subjectivity causes vagueness and inconsistency. Digital Signature contains all the information of objects in an image. Distance measurement between a query image and images in a database is based on cosine similarity coefficient. The result of our retrieval system is presented.

Important Links:

Go Back