Improving Land Use and Vegetation Cover Classification Accuracy using Fuzzy Logic - A Study in Pilibhit District, Uttar Pradesh, India

  • Nayak, S., and Beher, M.D. S


This study discusses a methodology for fuzzy classification and investigates the improvement of land use and vegetation cover classification accuracy of a fuzzy based classification over unsupervised classification in a test site of Pilibhit district, Uttar Pradesh, India. It also establishes a methodology of using higher resolution satellite data to gather reference points for classification accuracy assessment of outputs derived from low resolution images; apparently substituting ground validation. In this study the overall classification accuracy and overall kappa (K^) statistics of LUVC using fuzzy-based classification are 88.89% and 0.87 respectively, while the respective values are 82.22 and 0.79 of the unsupervised classification. This meets the requirement that K^ values >0.80 (i.e., >80%) represent strong agreement or accuracy between the classification map and the ground reference information, K^ values between 0.40 and 0.80 (i.e., 40 to 80%) represent moderate agreement, K^ values <0.40 (i.e., <40%) represent poor agreement (Landis and Koch, 1977 and Jensen, 2005). The study holds two promises i.e., (i) improved classification using fuzzy membership and (ii) utilization of higher spatial resolution data (merged data) to generate reference frame for evaluating classification accuracy.