The increasing role of agriculture in sustainable resource management has led to advancements in cropland mapping using remote sensing technologies and classification techniques such as Deep Learning and Object-Based Image Analysis (OBIA). While outdoor cropland classification has been well-studied, identifying crops under Plastic Covered Greenhouses (PCG) remains challenging due to the semi-transparent nature of greenhouse plastics. In 2015 and 2018, the authors of this work addressed the crop classification under PCG using OBIA, decision tree classifiers, and multi-temporal satellite imagery (Sentinel 2 and Landsat 8). In the 2015 work, a value of overall accuracy (OA) of 81.3% was reached for detecting four crops in Almería, with strong results for tomato and pepper but weaker performance for cucumber and aubergine. In 2018, we classified crops under PCG, obtaining OA values between 72.96% and 74.42%, with pepper being the most accurately classified crop. To build on these promising results, this study employs the same dataset, used in 2018 and corresponding to Autum 2016, to evaluate several classifiers, including decision trees, support vector machines, discriminant analysis, logistic regression, nearest neighbors, kernel approximation, naive Bayes, ensembles, and neural networks. The best results were achieved with the Linear Discrimination classifier, reaching 85% OA and a 0.78 kappa coefficient. Pepper and tomato showed the highest classification accuracy (F1-score values of 94.39 and 84.55, respectively).