Anggoro, Yerry (2017) Implementasi Metode Fuzzy K-Nearest Neighbor Untuk Klasifikasi Penyakit Tanaman Kedelai Pada Citra Daun. Sarjana thesis, Universitas Brawijaya.
Abstract
Kebutuhan protein sangatlah penting bagi tubuh manusia, salah satunya adalah kedelai yang merupakan sumber protein nabati. Selain jagung dan beras, kedelai merupakan komoditi pangan utama di Indonesia. Akan tetapi produksi kedelai dalam negeri belum memenuhi permintaan secara baik. Produksi kedelai di tingkat petani sebenarnya masih bisa ditingkatkan melalui inovasi teknologi, salah satunya yaitu mendeteksi penyakit tanaman kedelai pada daun dengan metode Fuzzy K-Nearest Neighbor dan segmentasi menggunakan metode Otsu. Citra diproses dengan metode Otsu untuk memisahkan bagian yang berpenyakit dengan bagian yang tidak berpenyakit lalu melakukan pengklasifikasian dengan metode Fuzzy K-Nearest Neighbor untuk menentukan penyakit karat daun, Downy Mildew, dan pustul bakteri. Terdapat empat pengujian yaitu pengujian perbandingan data latih dan data uji dengan akurasi tertinggi pada perbandingan 90:10 dengan jumlah 54 data latih dan 6 data uji sebesar 100%, pengujian terhadap nilai Threshold dengan T=10 menghasilkan akurasi 83,33%, pengujian terhadap nilai k=5 menghasilkan akurasi 83,33%, dan pengujian terhadap nilai m=2 dengan akurasi 83,33%.
English Abstract
Protein is one of essential thing to the human body, there are many source of protein and one of it is a soy which is nabati protein source. besides corn and rice, soy is the main food commodities in Indonesia. However, domestic production of soybean has not been enough to fulfil the necessity. Soybean production at the level of actual farmers could still be enhanced through technological innovation, one of that is to detect plant disease of soybeans on the leaves by the method of Fuzzy K-Nearest Neighbor and the segmentation using the method of Otsu. The image processed with the method of Otsu to separate parts that are diseased with parts that are not diseased and then do a classification by the method of Fuzzy K-Nearest Neighbor to determine leaf rust disease, Downy Mildew, and bacterial pustule. There are four tests such as test comparison data training and test data with the highest accuracy in comparison with a total of 90:10 54 training data and test data of 6 100%, testing against the values of Threshold with T = 10 generates 83,33% accuracy, testing against the values of k = 5 generates 83,33%, accuracy and testing against the values m = 2 with accuracy of 83,33%.
| Item Type: | Thesis (Sarjana) |
|---|---|
| Identification Number: | SKR/FTIK/2017/562/051708074 |
| Uncontrolled Keywords: | Otsu, Fuzzy K-Nearest Neighbor, kedelai |
| Subjects: | 000 Computer science, information and general works > 006 Special computer methods > 006.3 Artificial intelligence > 006.31 Machine learning |
| Divisions: | Fakultas Ilmu Komputer > Teknik Informatika |
| Depositing User: | Yusuf Dwi N. |
| Date Deposited: | 03 Oct 2017 04:21 |
| Last Modified: | 02 Sep 2020 07:23 |
| URI: | http://repository.ub.ac.id/id/eprint/3179 |
Preview |
Text
Bagian Depan.pdf Download (8MB) | Preview |
Preview |
Text
BAB I.pdf Download (5MB) | Preview |
Preview |
Text
BAB II.pdf Download (10MB) | Preview |
Preview |
Text
BAB III.pdf Download (5MB) | Preview |
Preview |
Text
BAB IV.pdf Download (17MB) | Preview |
Preview |
Text
BAB V.pdf Download (10MB) | Preview |
Preview |
Text
BAB VI.pdf Download (6MB) | Preview |
Preview |
Text
BAB VII.pdf Download (4MB) | Preview |
Preview |
Text
Daftar Pustaka.pdf Download (4MB) | Preview |
Actions (login required)
 |
View Item |