Supervised classification of crops and crop identification are in the tillage domain. This is a crucial topic to discuss. Crop feature extraction and crop categorization benefit greatly from hyperspectral remote sensing data. Hyperspectral data is unstructured, and deep learning methods work well with unstructured data. In this research, customized 3dimensional-Convolutional Neural Network (3D-CNN) was used as a deep learning method and the airborne visible/infrared imaging spectrometer sensor provided a standard dataset of Indian pines. And a study area dataset obtained from the airborne visible/infrared imaging spectrometer-next generation sensor for the extraction of crop features and crop classification. The case study took place at International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), Telangana India. The experiment shows that the customized 3D-CNN method achieves a good result with overall classification. The overall accuracy of the Indian pines dataset is 99.8% and the study area dataset is 99.5% compared to other cutting-edge methods.