SCALAR QUANTIZATION BASED LARGE OBJECT CATEGORIES FROM VISUAL SEARCH.

Content based image retrieval is a process. To retrieve a image stored in a database. That reinforces the mutual exchange of information into multiple modalities for improving search performance. It matching among images .The similarity measures are applied and evaluated in the content of approximate the original image detection. The proposed method uses SIFT (scalable invariant feature transform) quantized local feature descriptors. Problem of relies scalable visual search matching in large scale image search. It used method for bag-of-visual words (BOW).the method can be applied for image classification.BOW model; an image can be treated as a document, EX. words in image needed to define. Cascaded scalar quantization (CSQ) method is an ideal for more relevant images.