Edge Texture Based CBIR using Row Mean of Transformed Column Gradient Image

The paper discusses novel image retrieval methods based on edge texture of images extracted using gradient operators and slope magnitude technique with image transforms. Twenty eight variations of proposed image retrieval techniques using seven image transforms and four gradient operators like Roberts, Sobel, Prewitt and Canny are considered here. The proposed image retrieval techniques are tested on generic image database with 1000 images spread across 11 categories. In all 55 queries (5 from each category) are fired on the image database. The average precision and recall of all queries are computed and considered for performance analysis. The various „Mask-Shape-Transform‟ CBIR techniques [1] are compared with each other and each of the proposed „MaskShape-Transform‟ CBIR methods. The „Mask-ShapeTransform‟ is found to be better than the „Mask-Shape‟ Technique. In all Roberts-Kekre-Transform based CBIR gives best performance followed by Sobel-Slant-Transform and Sobel-Hartley-Transform.