Shape Based Image Retrieval Approach Using Complex Polar Cosine Transform

Abstract

The distinct features have dominant role in image retrieval domain, therefore, in this work we propose an accurate feature extraction method based on complex polar harmonic transform which conveys important information of the image. Bray-cuts and developed Bray- cuts distance measures are utilized to differentiate between the query and database images. The extensive experiments which carried out over three standards databases refer that the proposed method outperform polar cosine transform in image retrieval domain, also it is effective and robust against the noise. Furthermore it is rotation invariant.