OBTAINING A THRESHOLD FOR PARTITIONING A DATASET BASED ON CLASS VARIANCE AND CONTRAST

摘要

A method is proposed for obtaining a threshold value T which partitions a dataset, such as a medical image, into two classes, the classes being composed of pixels having an intensity value respectively not greater than and below the threshold. The threshold T is selected so as to minimize a weighted sum of (i) a term which varies with the within-class variance of the classes and (ii) a term which varies inversely with the contrast between the classes. The weighed sum depends upon a weighting factor λ which can be selected based on a supervised learning process using similar datasets, or obtained using a priori knowledge of the dataset.