PERBANDINGAN PEWARNAAN CITRA GRAYSCALE MENGGUNAKAN METODE K-MEANS CLUSTERING DAN AGGLOMERATIVE HIERARCHICAL CLUSTERING83

software that serves to draw. This jobs are very expensive and takes a lot of time. This study aims to improve the quality of grayscale images so that the results obtained have a better quality than the initial image, implementing techniques to improve image quality by creating a colorless image (gr...

Full description

Bibliographic Details
Main Authors: , Muhammad Safrizal, , Drs. Agus Harjoko,M.Sc, Ph.D
Format: Thesis
Published: [Yogyakarta] : Universitas Gadjah Mada 2012
Subjects:
ETD
Description
Summary:software that serves to draw. This jobs are very expensive and takes a lot of time. This study aims to improve the quality of grayscale images so that the results obtained have a better quality than the initial image, implementing techniques to improve image quality by creating a colorless image (grayscale image) into a color image, build a system that can be utilized to improve the quality of grayscale images, and compare of k-means clustering method and hierarchical agglomerative clustering methods for coloring grayscale images. Expected to be the basis of scientific development and implementation of image processing systems for improving the quality of grayscale image into a more complex color images. color The image is converted into the � ! color spaces. Luminance �! and gray value of each image are grouped based on proximity of each element. Color mapping is done by transferring the value of chrome !! to be added in the gray value of each element grayscale image with similarity between each centroid color images and. The results obtained in this study is a grayscale image color depends on the amount of a given cluster, the selection of the reference image categories and the influence of each color image in the same category. Testing staining based on subjective analysis and objective analysis that has been done can be concluded that the application of k-means clustering method is better than the application of agglomerative hierarchical clustering methods in coloring grayscale images.