The researchers used drone images to analyze vegetation in agricultural land and classify land use types. They calculated different vegetation indices like NDVI and found that NDVI was the most effective for monitoring vegetation. By using statistics like average and standard deviation of NDVI values, they were able to classify land use types like rice fields and corn fields accurately. The comparison with field data showed a high accuracy level of 0.914, indicating that using drone-based vegetation indices is effective for classifying simple land use types in areas like reclaimed farmland.