KLASIFIKASI CITRA DENGAN MATRIKS KO-OKURENSI ARAS KEABUAN (Gray Level Co-occurrence Matrix-GLCM) PADA LIMA KELAS BIJI-BIJIAN

Ganis K, Yudhistira and Santoso, Imam and Isnanto, R.Rizal (2011) KLASIFIKASI CITRA DENGAN MATRIKS KO-OKURENSI ARAS KEABUAN (Gray Level Co-occurrence Matrix-GLCM) PADA LIMA KELAS BIJI-BIJIAN. Undergraduate thesis, Jurusan Teknik Elektro Fakultas Teknik Undip.

[img]
Preview
PDF - Published Version
158Kb

Abstract

Human has an ability to classify image into appropriate classes. This ability will be very useful in many things if applied into a certain system which can be hardware or software. So it is necessary to perform a research concerning how far this ability of human being could be applied into a certain system. The goal of the research is developing software which have an ability to classify image into appropriate classes by using texture analysis, in this case is the image of seeds. The image of seed first transformed into a grayscale image, afterwards feature extraction is carried out. The method used to extract the feature is gray level co-occurrence matrix (GLCM). The features obtained from GLCM are energy, entropy, homogeneity, contras, correlation, inverse difference momentum, sum variance, sum average, sum entropy, difference variance, and difference entropy. The next step after feature extraction is classification by using k-nearest neighbor. The testing step uses six scenarios to find out the level of recognition of the software to the image of seeds. The testing results show that the highest level of recognition is the testing by using scenario 1, which is 100%, and the lowest level of recognition is the testing by using scenario 4 with close captured distance, which is 20%. The rice has the highest level of recognition, and the corn has the lowest level of recognition. Keywords : Feature extraction, co-occurrence, classification , k-Nearest Neighbor, seeds.

Item Type:Thesis (Undergraduate)
Subjects:T Technology > TK Electrical engineering. Electronics Nuclear engineering
Divisions:Faculty of Engineering > Department of Electrical Engineering
Faculty of Engineering > Department of Electrical Engineering
ID Code:25431
Deposited By:Mr. Sudjadi Pranoto
Deposited On:12 Jan 2011 14:13
Last Modified:12 Jan 2011 14:13

Repository Staff Only: item control page