Otsu's Implementation Otsu's method works fine whenever an image has two classes associated with it. However the Otsu's method must be modified in order to decide the appropriate value for threshold for an image that has more than two classes. It is unsupervised. Also the author mentions that as the number of classes increases the selected thresholds decrease. The Otsu's method is used to minimize the intra-class variance and thus it can be thought of as a brute force method that enables searching and selecting the most optimal threshold for an image. It can be termed as the most standard and easy to understand method for selecting threshold automatically and also can be applied to several problems.Apart from that it has implementations in MATLAB.
How to run the code:
- Open MATLAB and select the filepath where fn_Otsu.m is stored.
- Give filename as 'example.jpg'.
- Run your program.