We use a sequence from the Aqu@theque dataset.
Fig. 1. First row: The current image, the ground truth. Second row: Sugeno-Ohta, Choquet-Ohta. Third row: Choquet-HSV and Choquet-YCrCb.