CBIR using Upper Six FFT Sectors of Color Images for Feature Vector Generation

In this paper we are using Fast Fourier Transform to generate the feature vector which considers the mean real and mean imaginary parts of complex numbers of polar coordinates in frequency domain. The method proposed here considers 12 mean values of 6 upper half sectors real and imaginary parts of each R, G and B components of an image. The algorithm proposed uses 36 mean values of real and imaginary parts in total. The proposed work experimented over a database of 249 images spread across 10 classes of images. Euclidian distances between the feature vectors of query image and the database images are considered. Images are retrieved in ascending order of Euclidian distances. The Average precision and Average recall of each class and overall average of all averages of each class are calculated as a performance measure. The cross over point of average recall and precision is 50% and it is 40% or above for all classes.