Performance Analysis of Support Vector Machine, Neural Network and Maximum Likelihood in Land Use/Cover Mapping and GIS (A Case Study: Namin County)

The earth’s surface study by remote sensing has many benefits such as continuous acquisition of data, broad regional coverage, cost effective data, map accurate data, and large archives of historical data. Land use/ cover mapping in natural resources and environmental management and providing land use plans as well as determining land ability are necessary to create developing plans. Here a methodology has been development to map land use/cover using Landsat Operational Land Imager (OLI) data in Namin County of Iran for June 23, 2015. Because of OLI production with radiometric and geometric correction do not needs to correct them but the atmospheric correction apply with Quick method. FCC were used in this study are b5 from principle image as NIR band, NDVI that extracted from principle image as Red band and PCA 1 used as green Band. Based on the Anderson land-use/cover classification system, the land use/covers are classified as forest land, settlement, water bodies, range land, agricultural land and bare land. The land use/cover maps was produced by using supervised image classification technique based on Maximum Likelihood Classifier (MLC), Neural Network Classifier (NNC), support vector machine (SVM) with four kernels (linear, polynomial, radial basis function and sigmoid) and 261 training samples. Then, to remove the individual and distributed pixels in classified map and also to have a uniform map, 3×3 majority filter was implemented on the map. For accuracy assessment used from the Quick Bird satellite image was used and randomly collect 180 samples point as ground truth. Overall accuracy, user’s and producer’s accuracies, and the Kappa coefficient were then derived from the error matrices. The result showed that SVM (Linear & Radial) with kappa coefficient (0.989) and overall accuracy (99.1) % are the most accurate in order to produce land use/cover map.

Keywords:

MLC , NNC , SVM , OLI , Namin County , Iran