Content Based Image Retrieval using Color, Multi Dimensional Texture and Edge Orientation
J. Sammanasu Mary, S. Christina Magneta · International Journal For Science Technology And Engineering · 2016
Many image retrieval systems use global features to retrieve images such as color and texture. But the previous results provide too many false positives while using these global features to search for similar images. Content-based image retrieval system works with the entire image and searching is based on comparison of the query image with image database. CBIR uses the visual information of an image such as color, texture and edge to represent and index the image. These contents can be obtained by applying color, texture and edge based techniques. These techniques are applied on both query image and database to get an image from the database more accurately. Color images are indexed by error diffusion block truncation coding (EDBTC). To improve the accuracy of color histogram based matching YCbCr color space is added into EDBTC indexing scheme. Texture analysis is more important because it is used to improve the discriminatory ability of the extracted image features. The speed of edge based retrieval and texture based retrieval can be enhanced by using canny edge detection and gray level co-occurrence matrix. GLCM is used to extract second order statistical texture feature of image. These proposed techniques are used to improve the accuracy of the result.