An efficient image retrieval through DC feature extraction

We propose a new simple method of DC feature extraction that enables to speed up and decrease storage needed in image retrieving process by aim of partial Joint Photographic Experts Group (JPEG) compressed images algorithm. Our feature extraction is carried out directly from JPEG compressed images. We extract one component of DCT coefficients which is DC component only. By applying DC feature, we only need 1/64 times that is required to construct indexing keys compared to conventional Euclidean distance method whilst maintaining high precision and recall. We also use a DCT coding category approach to compare our proposed method. After all, even though this method is not revolutionary and perfect yet in improving speed for image indexing, it reduces significantly the cost of image retrieval in real applications, primarily for huge image databases.