| Summary: | 碩士 === 銘傳大學 === 應用統計與資料科學學系碩士班 === 105 === In this research, we used Nakagami-model-based image to identify the region of interest (ROI) and use the estimation to estimate the Nakagami parameter, which can be estimated in different methods. To avoid the errors in calculation, we used the maximum likelihood approach, supposing the selected samples which are the most likelihood and the estimation which was calculated by the population of the density function, and the moment estimation. Furthermore, we considered the relevance of each variable, so we performed the Weibull, gamma, chi-square, Rayleigh, and Nakagami. Last but not least, we assumed the maximum likelihood estimation is more precise than the moment estimation in examining the population standard deviation of the Nakagami parameter.
|
|---|
