A literature survey for fuzzy based soft classification techniques and uncertainty estimation

Remote sensing is a popular and fast way to acquire information on geographic surface. The multispectral images are main source of information on geo-surface. Classification is used to obtain this data and constructing maps. In general, the classification is crisp and assigns a single class to each image elements or pixel, known as hard classification. However, there is another consequence that some pixels may not purely belongs to single class. This problem treated as mixed pixel problem. Mixed pixels are big issue in term of hard or static classification. Due to hard classification the significant amount of result is lost. but we can bring under control this problem to using soft classification. Classification is not completed without assessment of accuracy. Accuracy assessment of classification play vital role in term of classification and it tells amount of accuracy for the classification process. Accuracy assessment is an important issue for validation of image classification. There are several tools that provide accuracy analysis using confusion matrix, Kappa coefficients. This assessment is based on the rule of single class per pixel known as hard classification. Further for Mixed pixels in classification handled by soft classification i.e. assigning membership of each class to that mixed pixel and accuracy assessment is also differ from hard classification as each pixel have membership proportion for every class which involved in classification schema. Further assessing accuracy also depends upon how samples have been taken from census of classified output. This paper study is to focus on comparative study in soft classification approaches and accuracy assessment with reference data and without reference data pixels.