New Feature Descriptors for Content Based Image Retrieval. Development of New Feature Descriptors for Content Based Image Retrieval

      Аннотация: The proliferation of high quality and economical image sensors has precipitated renewed interest in the research community for developing efficient techniques to handle a large database. The concept of image retrieval came into existence and has been an active area of research aimed at addressing two major issues viz., database management and computer vision.The visual content descriptors are either global or local. A global descriptor represents the visual features of the whole image, whereas a local descriptor represents the visual features of regions or objects to describe the image. Further, these are subdivided into two categories, spatial and transform domain-based features. The approach makes use of pixels (a group of adjacent pixels) or gray values and the other makes use of transformed data of the gray image for feature calculation.

Ключевые слова: color, texture, precision, recall, similarity index, corel, broadtz, MITvistex, LBP, LMEBP, DBC, Cooccurance matrix, rotation invariant, histogram