Brain Image Decomposition on Image Alteration

Abstract

Image transformation is essential to explore and find out specific information that does not exist and has not been previously known from an image, such as pixels, geometry, size or colour. Therefore, this paper aims to analyze the image transformation by generating the value of the thresholding method from low to high in image segmentation. The segmentation process works based on two-colour models, namely HSV and RGB colours. Image segmentation problems occur when an object has more than two colours. There is lighting that affects the condition of the thing, so it is proposed to change the threshold value from low to high to compare the segmentation results of each given value to determine the optimal threshold value. The image data used is a brain MRI image from the website www.hipwee.com. The Matlab 2019a application is used to assist the analysis process. Based on the analysis results, in this case, the smaller the HSV and RGB threshold values, the segmentation of the detected image results is not detailed and the less precise, and vice versa. The best Brain MRI Image Segmentation results when using a threshold value of HSV=0.5 and RGB=5.