Image Retrieval Based Upon Compressed Camera Identification Using NN Classification and Regression Model

Image retrieval is a method to retrieve digital images. Now these days retrieving images from large collection of database according to specific criterion is an important task. An important criterion is the retrieval of images based on specific camera. This research paper focus on a well established fingerprint matching technique to link pictures to the specific camera that acquired them and for the fast retrieval images are compressed that clicked by different cameras. High retrieval efficiency and less computational complexity are desired characteristics of image retrieval systems. In this paper an image retrieval system is proposed which will provide accurate results as compared to previous developed system. In this paper SVM (Support vector machine), KNN (knearest neighbor) and NN (Neural network) classifier is used and GLCM (Gray-level co-occurrence matrix) is for retrieve the contents of the images based on color, Texture etc. The precision and recall values calculated according to the retrieved contents of the images from the datasets.