Review on Classification of Natural and Synthetic Images

The classification of image system mainly focuses on photographic and nonphotographic i.e. natural and synthetic images. The semantic description-based on classification of images is a more interesting and significant problem in involuntary image identification. An algorithm for natural and synthetic image classification system has been established. In order to adventure the difference, the color pattern and spatial correlation of pixels in natural and synthetic images, some of the features are removed from the database of images. If these features are used alone, they have low accuracy but when it combined together and used for image classification forms a more multifaceted and appropriate classification and their exactness can be improved. Proposed image classification for natural and synthetic images algorithm will be use these low-level features of images such as edge map, color map, threshold ratio, nearest neighborhood and energy level for classification of the images into synthetic and natural.