Propose a region-growing algorithm to segment the image in Problem 10.36.10.36 The objects...

Propose a region-growing algorithm to segment the image in Problem 10.36.

10.36 The objects and background in the image shown have a mean intensity of 170 and 60, respectively, on a [0,255] scale. The image is corrupted by Gaussian noise with 0 mean and a standard deviation of 10 intensity levels. Propose a thresholding method capable of yielding a correct segmentation rate of 90% or higher. (Recall that 99.7% of the area of a Gaussian curve lies in a ±3s interval about the mean, where s is the standard deviation.)