Content Based Image Retrieval on Image Sub-blocks
Ayisha Begam, K. A. Abdul Nazeer · 2013
This paper proposes a content primarily based image retrieval (CBIR) system exploitation the local color and texture options of chosen image sub-blocks and world color and form options of the image. The image sub-blocks are roughly known by segmenting the image into partitions of different configuration, finding the sting density in every partition exploitation edge thresholding, morphological dilation and finding the corner density in every partition. The color and texture options of the known regions are computed from the histograms of the quantized HSV colour area and grey Level Co- incidence Matrix (GLCM) respectively. A combined color and texture feature vector is computed for every region. the form options are computed from the sting bar graph Descriptor (EHD). Euclidian distance live is employed for computing the distance between the options of the question and target image. Experimental results show that the planned methodology provides higher retrieving result than retrieval exploitation a number of the present strategies.