SEGMENTATION-BASED WINE IMAGE IDENTIFICATION SYSTEM USING THE K-MEANS CLUSTERING METHOD

Abstract

Computerized Image Handling is a field of science that concentrates on how images are framed, handled, and examined to produce data that people can understand. Image sharing interactions can be solved by applying existing techniques. Many techniques can be used, for example the Gabor Channel strategy, GLCM technique, Wavelet strategy, Area Development strategy, K-implies Grouping strategy, Mean Shift Bunching strategy. In this test, we tried the process of dividing image variations in pictures of grapes. Division is an important part of image inspection, because in this system the desired image will be dissected for further processing to make it easier to break down, for example on design ID. The image division can be partitioned into several stages in the cycle of examining and recovering several objects of interest. One strategy for dividing images is grouping. Bunching is the work of grouping information based on classes and is a strategy for collecting information in a dataset. The division of images and differentiation of evidence in this exploration uses the K-Means clustering technique. K-Means is a simple and fast computational strategy, before cutting or differentiating leaves, first determine the variation space using CIELab. ID testing information uses two methodologies, namely shape inspection and surface investigation.