CONTENT BASED IMAGE RETRIEVAL BERBASIS COLOR HISTOGRAM UNTUK PENGKLASIFIKASIAN IKAN KOI JENIS KOHAKU

SYARIF, HISYAM (2014) CONTENT BASED IMAGE RETRIEVAL BERBASIS COLOR HISTOGRAM UNTUK PENGKLASIFIKASIAN IKAN KOI JENIS KOHAKU. Skripsi, Fakultas Ilmu Komputer.

Preview

PDF - Published Version Download (4Kb) | Preview

Abstract

In recent years the collection and processing of data in the form of images is growing very rapidly. A data in the form of large amounts of imagery used in various fields, one of them in the field of fisheries, especially for the determination or classification of types of koi fish. Lately koi fish are very popular among the public to serve as decoration to beautify the house, or just a hobby to maintain. But some fans do not know the type of koi koi what possessed him, only interested in the size and color of its unique and diverse. Types of koi can be distinguished from the patterns in body color, the introduction of color in koi problem can be solved by applying the method of Content Based Image Retrieval based on Color Histogram with Euclidean Distance and Mean Square Error to get the error level of accuracy. By applying this method, based on the determination of the type of koi fish koi shades of color on the body can be resolved. CBIR is the application of computer vision to the image search techniques are taken from a data base that provides image. Based on the results of tests performed on 15 different Koi types using Euclidean Distance, found 3 types of Koi that are similar to the reference image Kohaku1, Kohaku2 and Kohaku3 the threshold value between 0-70000000 that indicates if the image is similar. And to get the level of accuracy of the techniques used his error Mean Square Error with the results of the threshold between 0-213000, which indicates if the image is similar. Mean Square Error of the results obtained validate the results of 330 931 pixels for image selection accuracy

Actions (login required)