In remote sensing applications, there exist numerous challenges in the classification of Hyperspectral images due to large amount of information processing which is required for classification, and the scarcity of labeled samples. Identifying the crops automatically is a key application in the Hyperspectral image analysis. In this article, we analyze the Support Vector Classification (SVC), Deep Neural Network (DNN) and Fusion Spectral Convolutional Neural Network (FuseNET) for crop cover classification. The results obtained for each of these methods, on standard benchmark datasets such as Indian Pines (IP), Pavia University (PU), Salinas (SA) and WHU-Hi-Longkou (WU), are presented and analyzed. The results demonstrate that FuseNET demonstrates an improvement of 13.81% for IP, 4.70% for PU, 7.13% for SA and 3.1967% for WU when compared to DNN. Also, FuseNET shows an improvement of 27.99% for IP, 18.22% for PU, 15.166% for SA and 8.02% for WU when compared to SVC