Susanto, Irwan Dwi Wahyu (2016) Perancangan Sistem Uji Ketahanan Kualitas Warna Akibat Gosokan Pada Bahan Kulit Berbasis Android. Undergraduate thesis, Institut Teknologi Sepuluh Nopember Surabaya.
Preview |
Text
2414105008-Undergraduate_Thesis.pdf - Published Version Download (1MB) | Preview |
Abstract
Pengujian bahan kulit menurut SNI ISO 105-A02:2010
dilakukan dengan pengamatan visual. Pengujian ini membutuhkan
waktu lama dan melibatkan minimal 3 orang pengamat. Agar
pengamatan menjadi subyektif perlu adanya inovasi teknologi yang
mampu menggantikan pengamatan visual, yaitu aplikasi yang dapat
digunakan untuk membantu melakukan pengujian kualitas bahan
kulit berbasis android. Bahan kulit yang digunakan dalam pengujian
warna diuji gosok basah dan kering kemudian diidentifikasi
berdasarkan pengolahan citra. Tahap pengolahan citra, yaitu tahap
preprocessing dilakukan dengan melakukan proses cropping dan
resize citra menjadi 50x50 piksel, tahap ekstraksi ciri menggunakan
ekstraksi ciri statistik orde pertama (mean, standar deviasi, variance,
skewness dan kurtosis) dan tahap klasifikasi menggunakan metode
K-Nearest Neighbour (KNN) berdasarkan perhitungan jarak
euclidean. Hasil pengukuran menggunakan metode ekstraksi ciri
orde pertama memiliki akurasi tertinggi dengan ciri mean sebesar
95,454% pada uji gosok basah dan 87,272% pada uji gosok kering
menggunakan metode KNN dengan K (tetangga terdekat) = 1
============================================================================================
Leather materials testing according to ISO 105-A02: 2010
conducted by visual observation. This test takes a long time and
involve at least three people watchers. In order to be subjective
observations need for innovative technology that can replace visual
observation, which is an application that can be used to assist with
testing the quality of the leather-based android. Materials used in
the test skin color tested wet and dry rub then identified based on
image processing. Stage image processing, the preprocessing stage
is done by performing the process cropping and resizing images to
50x50 pixels, stage of feature extraction using feature extraction
statistics first order (mean, standard deviation, variance, skewness
and kurtosis) and stage classification using K-Nearest Neighbour (
KNN) based on euclidean distance calculations. The measurement
results using feature extraction method first order has the highest
accuracy with the characteristics of a mean of 95.454% in the wet
rub test and 87.272% on a dry rub test using KNN with K (nearest
neighbor) = 1
| Item Type: | Thesis (Undergraduate) |
|---|---|
| Additional Information: | RSF 005.43 Sus p |
| Uncontrolled Keywords: | kulit uji, skala standar grayscale, prepocessing, ekstraksi ciri, klasifikasi |
| Subjects: | Q Science > QA Mathematics > QA76.774.A53 Android |
| Divisions: | Faculty of Industrial Technology > Physics Engineering > 30201-(S1) Undergraduate Thesis |
| Depositing User: | EKO BUDI RAHARJO |
| Date Deposited: | 02 Jul 2020 04:17 |
| Last Modified: | 02 Jul 2020 04:17 |
| URI: | http://repository.its.ac.id/id/eprint/76283 |
Actions (login required)
 |
View Item |