BW Trained HMM based Aerial Image Segmentation
Image segmentation is an essential preprocessing tread in a complicated and composite image dealing out algorithm. In segmenting arial image the expenditure of misclassification could depend on the factual group of pupils. In this paper, aggravated by modern advances in contraption erudition conjecture, I introduce a modus operandi to make light of the misclassification expenditure with class-dependent expenditure. The procedure assumes the hidden Markov model (HMM) which has been popularly used for image segmentation in recent years. We represent all feasible HMM based segmenters (or classifiers) as a set of points in the beneficiary operating characteristic (ROC) space. optimizing HMM parameters is still an important and challenging work in automatic image segmentation research area. Usually the Baum-Welch (B-W) Algorithm is used to calculate the HMM model parameters. However, the B-W algorithm uses an initial random guess of the parameters, therefore after convergence the output tends to be close to this initial value of the algorithm, which is not necessarily the global optimum of the model parameters. In this project, a Adaptive Baum-Welch (GA-BW) is proposed.
Keywords: Convex hull, hidden Markov models, image segmentation, ROC convex analysis, ROC curve, Genetic Algorithm; HMM training; Baum-Welch algorithm
Download Full-Text
