On-the-Go Wheat Head Counting Using Robotics and Deep Learning Datasets

Navigation: This dataset includes the original images captured in the field by the camera mounted on the wheel of the Thorvald robot, along with annotated images featuring masks and lines. It also contains the training, validation, and test split, as well as text files with the predicted values for the line after training the model. Additionally, the dataset includes model checkpoints and weights for the trained model.
Wheat Head Counting: This dataset consists of images captured in the field by the camera with a top-down view mounted on the Thorvald robot, which were used for wheat head counting and the results of the counting model for each plot and resolution.