Classifiers in Cascade as a way to distinguish the desired shape. By bringing such recognition and repeat multiple revolutions, in which each round, it will cut the image in Negative space in every round that is found. Then at the end of the process will be the number of Sub Window that is Negative, there will be reduced until the desired shape and recognition in order to find the desired shape, this is done with the image reflected in the thikran topic 2 to achieve speed in processing, in which the steps of the Cascade Classifiers can be shown as: illustration 2.8.
การแปล กรุณารอสักครู่..