Amrullah, Syakir Almas (2017) Perancangan Sistem Inspeksi Visual Berbasis Computer Vision untuk Penggolongan Buah Apel. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
2412100077-Undergraduate_Theses.pdf Download (2MB) | Preview |
Abstract
Buah apel memiliki beberapa macam varian yang masing-masing memiliki karakter khas yang berbeda. Perbedaan karakter antar varian menjadi penting untuk diketahui karena berkaitan dengan banyak hal dari sifat tanam dan kandungan kimia, hingga nilai jual dan pemanfaatannya. Proses inspeksi yang dilakukan secara manual dalam pengenalan jenis varian memiliki beberapa kekurangan seperti subjektifitas, inkonsistensi, tingkat kejenuhan, ketergantungan pada pengalaman. Penelitian ini bertujuan untuk merancang dan menerapkan teknik computer vision pada sistem pengenalan varietas buah apel berdasarkan penciri-penciri visual khas varietas buah apel. Citra buah apel yang didapat menggunakan webcam disegmentasi dengan metode Canny. Kanal warna hue dan vektor kontur diekstrak dari citra untuk kemudian disimpan sebagai data acuan dan dibandingkan dengan nilai pada data uji. Proses pembandingan kedua penciri tersebut dilakukan dengan membandingkan selisih histogram kanal hue dari citra uji dengan citra acuan. Sedangkan kemiripan bentuk antara citra uji dengan acuan dilakukan dengan membandingkan kontur dari citra uji dan citra acuan. Proses pengambilan keputusan dari proses pengenalan ini dilakukan dengan menggunakan algoritma K-Nearest Neighbor.
==================================================================================================
Apple fruit has many varieties and each variety can be identified by its unique features. Recognizing apple variety becoming important because those are different in cultural behaviour, chemical characteristics, price, and utility. Manual labour based inspection proces tend to have several drawbacks in recognizing diffetent varieties, such as its subjectivity, inconsistency, saturation level, and experience-dependent. This work goal is to design and implement computer vision technique in apple variety recognition system using its color and shape feature. Apple image which captured using webcam is segmented using Canny method. Hue color channel and contour vector are extracted from captured image and stored as database to be compared with new data then. Comparation of color features determined by comparing hue channel histogram of new data and those from database. In other hand, to determine matching level of tested data shape to database, both Hu moment value is compared. Decision making process in this work is based on K-Nearest Neighbor algorithm.
| Item Type: | Thesis (Undergraduate) |
|---|---|
| Uncontrolled Keywords: | computer vision, HSV color space, Hu moments, KNN |
| Subjects: | S Agriculture > S Agriculture (General) T Technology > T Technology (General) > T58.62 Decision support systems T Technology > TA Engineering (General). Civil engineering (General) > TA1637 Image processing--Digital techniques |
| Divisions: | Faculty of Industrial Technology > Physics Engineering > 30201-(S1) Undergraduate Thesis |
| Depositing User: | SYAKIR ALMAS AMRULLAH |
| Date Deposited: | 17 Mar 2017 03:59 |
| Last Modified: | 17 Mar 2017 03:59 |
| URI: | http://repository.its.ac.id/id/eprint/41194 |
Actions (login required)
 |
View Item |