Crop type mapping using hyperspectral remote sensing images has become increasingly important for analyzing agricultural landscapes. Convolutional neural networks (CNNs) are widely employed for this task by virtue of exceptional efficacy in processing visual data. In this study, we propose a 2D CNN, 3D CNN along with other algorithms such as naive Bayes, decision tree, and random forest for hyperspectral image classification. The 3D CNN mechanism, known for its robustness in feature extraction from spatial data, is particularly suited for this task. In our evaluation, we assess the effectiveness of our approach by employing three well-established remote sensing datasets: the Kennedy Space Centre dataset, the University of Pavia Centre dataset, and the Salinas SceneA dataset. These datasets, across multiple bands, provide rich data for crop type classification. Our experimental findings underscore the efficacy of the suggested approach in precisely categorizing crop varieties from hyperspectral remote sensing imagery. The comparative analysis with traditional machine learning algorithms showcases the superiority of the 3D CNN approach in handling complex spectral data and achieving higher classification accuracy. This research contributes to the advancement of agricultural monitoring and management by providing a reliable method for automated crop type mapping using remote sensing technology.