A Non Linear Object Classification using Object Based Approach

Remotely sensed imagery’s resolution has risen significantly in both the spectral and spatial domains during the past few years, providing a large amount of spectral and spatial data. This paper discusses an object classification approach on incorporating K-Nearest Neighbour (KNN) algorithm for segmentation and classification of non-linear objects. Analyzing digital imagery is done by using object-based classification (OBC) technique. Both spectral and spatial information are used for classification. In this study various features are extracted from the spatial and spectral information of image objects for classification of non-linear objects. Spectral features of the training image objects are extracted using region of interest (ROI) based samples which are used in KNN algorithm for segmentation and classification and achieve a good level of accuracy. The primarily focus of this work is to extract the non-linear objects by avoiding misclassification in a compact manner and also to improve the visibility of object borders.