Shape Recognition and Segmentation Using Active Contour Model and EM Algorithm for MRI and Color Images

摘要

Recent advances in imaging systems and high performance computer graphics have created a unique opportunity to develop a novel set of application in fruit industry. MRI and color imaging systems have the potential to be an integral component in pre- and post harvest investigation of physiological changes in fruit and vegetables. To evaluate the surface quality, the shape of the pomegranate fruit should be recognized from background. Because of special shape of pomegranate, active contour model (ACM) is used to detect shape of fruit. Furthermore, MRI is used to visualize the internal structure of pomegranate fruit. For this purpose, we use EM (Expectation Maximization) algorithm to segment pomegranate fruit. The extracted features can be correlated with internal fruit quality.