Table 1.
Between-Class Variance | Parameters Values | Effectiveness |
---|---|---|
The Original Otsu | Mean and variance are based on the definition of Gaussian distribution | Suitable for images with symmetric distribution but limited for asymmetric distributions |
The Modified Otsu [3] | The original Otsu formula using only the mean value of the lognormal distribution | Almost have the same efficiency as the original method with improvements for certain types of images |
The Proposed Model | Mean and variance are based on the definition of Lognormal distribution | Boosted efficacy for images with right-skewed intensity distributions but not suitable for left-skewed distribution. |