A review of assessing the accuracy of classifications of remotely sensed data