Performance analysis of fuzzy convolution tactic for classification of spatial pattern and crop area

In this work we have used data as acquired by IRS P6 LISS III sensor to classify crop area and spatial pattern of Vaijapur Tehsil. The supervised classification approach with fuzzy convolution technique and Maximum likelihood classifier has been used for analysis of satellite image for the month of February 2015 and ground truth data collected during field visit for testing and training purpose. Fuzzy convolution has been applied on the fuzzy layers obtained after fuzzy classification and compared these result with Maximum Likelihood classifier. It is found that Maximum likelihood has given better result for overall classification but for crop area class; fuzzy convolution gives more truthful results.