A novel method of Image Segmentation-Analysis

In this paper, we present the methods of abstracting separation of satellite imagery; we used seven techniques at this s, Kiresh method, Laplacian technique, Canny strategies, Robert's strategies, and Edge Maximization Technique (EMT) is also compared to another to choose the best way to get the edge section image. These techniques are used on a single satellite image to select the backstage; Sobel operator technique, Prewitt techniques guess the partition or edge adoption image. We propose a novel multi-scale acquisition edge and a vector field design plan. We show that the use of multiscale techniques to obtain the edge and quality of separation in natural images can be greatly improved. Our approach eliminates the need to choose a clear scale and follow the edge. Our approach prefers existing edges over a wide range of scales and makes these edges geographically with the best scales. This work is then extended to separate images by multiple scales using our anisotropic distribution system.