Remotely Sensed Image Segmentation using Multiphase Level-Set ACM

In remote sensing image analysis, segmentation of an image is an important aspect. It classifies similar pixels within the image. Image Segmentation is helpful in analyzing the patterns, objects, and edges within an image. There are many ways for performing image segmentation. In this paper, we are segmenting a satellite image using Multiphase Chan-Vese model. Chan-Vese models are based on ‘Active Contours without edges’. Active contour model is also known as Snake and Energy-Based Model, which is finding local minima in the equivalent energy function. Chan-Vese model gives effective results of segmented image. The multiphase level set construction is mechanized to avoid the drawback of overlap and vacuum; it can also signify edges with convoluted topologies. Researchers conclude in this paper with the findings that the multiphase CV method can give a sensible segmented image of satellite imagery with 2D-DWT, when they manipulate Heaviside function.