An Exact and Fast Exhaustive search in Otsu Thresholding Method

In this paper a new algorithm to find the optimal threshold in Otsu thresholding method is proposed. Because of evaluating whole existing gray levels as candidates for optimal threshold, this algorithm is completely exact. Moreover, the algorithm is fast due to recursive performance. All parameters which are needed to calculate between-class variance for each gray level can be obtained in terms of previous gray level or next gray level. This reduces the computation cost efficiently.

کلیدواژه ها:

Thresholding Method ، image segmentation ، image classification ، histogram information