An Efficient Approach of Content Based Image Retrieval using Texture, Color and Shape Features of an Image

摘要

retrieval of an image is an extra powerful and green for dealing with enormous photograph database. content based photograph retrieval (CBIR) is a one of the picture retrieval technique which makes use of consumer visible capabilities of an photograph together with color, form, and texture functions and so forth. it permits the quit user to offer a query photograph so one can retrieve the saved photos in database in keeping with their similarity to the question photograph. on this work, content based totally image retrieval is done by means of combining the two functions including color and texture. Color capabilities are extracted by way of using HSV histogram, color correlogram and color second values. Texture capabilities are extracted through segmentation based fractal texture analysis (SFTA). The accuracy of color histogram based totally matching can be accelerated by way of the use of shade coherence vector (CCV) for successive refinement. The velocity of shape based retrieval may be enhanced by way of considering approximate form in place of the exact form. The principle goal this work is classification of photo the use of SVM algorithm.