AnnaSaro Vijendran - Director,Master of Computer Application,SNRSons College, Cbe, India
E S. Vinod Kumar - Assistant Profssor in PG computer Science,STC, Pollachi, India

Due to the event and improvement in web with high speed for the previous few years and also the availableness of an oversized digital image assortment, economical image retrieval systems square measure needed. Retrieval may be text-based and content-based.Content primarily based image retrieval (CBIR) may be a method of retrieve and show relevant pictures from giant assortment of image info on the idea of their visual content. CBIR is employed for retrieval of pictures relying upon visual contents of pictures referred to asoptions. This paper focuses on color and texture primarily based techniques for achieving economical and effective retrieval of pictures. Color feature extraction is finished by RGB to HSV conversion and color moments (mean and normal deviation). Texturefeature extraction is noninheritable by Gabor wave filter and wave moments (mean and normal deviation). For classification of extracted options we ve got used support vector machine (SVM). geometer distances square measure calculated of each options for similaritymeasures.

CBIR, Color feature, geometer distance,Gabor wave filter, RGB to HSV, SVM, Texture feature