Land cover classification of SSC image: unsupervised and supervised classification using ERDAS imagine

Abstract

NASA's Earth Sciences program is primarily focused on providing high quality data products to its science community. NASA also recognizes the need to increase its involvement with the general public, including areas of information and education. The main objective of this study is to classify the vegetation, man-made structures, and miscellaneous objects from the Satellite Image of NASA Stennis Space Center (SSC), Mississippi, and USA, by using the software, ERDAS Imagine 8.5. The ERDAS Image software performs the classification of an image for identification of terrestrial features based on the spectral analysis. For classification of the SSC image, the multispectral data was used for categorization of terrestrial objects, vegetation and shadows of the trees. These are two ways to classify pixels into different categories: Supervised and unsupervised. The classification of unsupervised data through ERDAS Image helped in identifying the terrestrial objects in the Study Image (SSC). The spectral pattern present within the data for each pixel was used as the numerical basis for categorization. The first analysis of the Image SSC involved the use of generalized Unsupervised Classification with 4 categories (Grass, Trees, Man-Made and Unknown). The result of the Unsupervised Image was used to create another image by using Supervised classification. The key difference between the two images is the ability of supervised image to decipher similar images, such as the roofs of the buildings and the shadows of the trees. Image stacking was conducted to create a fully classified image to separate shadow, grass, man-made, and trees. This poster will describe the procedures for viewing and measuring image, Computer-guided (Unsupervised) and User-guided (Supervised) Procedures will be described on image stacking to view each classification one at a time and stack them into a complete Classified Image. The application of unsupervised and supervised classification in agriculture will be discussed by giving examples of measurement of field reflectance of two classes of giant salvinia [green giant salvinia (green foliage) and senesced giant salvinia (mixture of green and brown foliage)], and invasive aquatic weed in Texas.