The Design and Implementation of Plant Disease Spot Segmentation Algorithm Based on Improved CV Model

Abstract

In view of the shortcomings of which CV model can’t divide grayscale uneven image and need to set the initial segmented contour manually and so on. This paper introduces Otsu algorithm based on the CV model, to make sure the Initial segmented contour and avoid the shortcomings which need to set the initial segmented contour manually, at the same time, making CV model get good segmentation effect on gray uneven image through introducing local variance. The result proves that though the number of iterations is less than original CV model, it also can get better effect than original CV model, Closer to the profile of plant disease in this paper, meanwhile, the method adopted by this paper make computing speed faster than the original algorithm.