Crop-type mapping and crop classification is becoming increasingly important for agricultural forecasting, economic, and governmental purposes, and with the rise of machine learning for image classification, the implementation of models for this purpose is becoming highly relevant for both public and private industries. Issues have existed though in the past with the size of datasets and the difficulty in obtaining accurately labeled crop-types; however, using the TinyEuroCrops dataset, which organizes the data in an easy-to-work-with HDF5 format and gives just enough data to accurately train a model, an image processing model was constructed to accurately ( \(> 85\%\) top 1 prediction accuracy) predict the crop-type of more than 340,000 parcels of land in Austria. This model integrates multilayer perceptrons, a multi-headed attention layer, and a temporal encoder to effectively classify 44 crop types across the country. The employed techniques are ubiquitous enough so as to be expanded to larger datasets with a variety of crop-types easily.