SARI, WINDA NOVITA and Praptini, Puji Catur Siswi and Haris, Abdul (2021) SMART COMPUTATION SYSTEM MODELS (SCSM) UNTUK IDENTIFIKASI PENYAKIT PADA DAUN CABAI MENGGUNAKAN ALGORITMA LEARNING VECTOR QUANTIZATION. Diploma thesis, ITPLN.
![[thumbnail of 201731349_ Winda Novita Sari_Revisi_Skripsi_winda novita sari.pdf]](https://repository.itpln.ac.id/style/images/fileicons/text.png)
201731349_ Winda Novita Sari_Revisi_Skripsi_winda novita sari.pdf
Restricted to Repository staff only
Download (4MB)
Abstract
Tanaman cabai merah merupakan kelompok sayuran yang diminati masyarakat Indonesia. Permasalahan yang sering terjadi adanya organisme pengganggu tanaman cabai sehingga dapat membuat produksi cabai menurun. Adanya penyakit tanaman cabai yang sulit dikenali oleh petani dengan menggunakan mata dan tanpa menggunakan alat bantu. Penyakit tanaman cabai memiliki ciri-ciri tertentu. Dalam menyelesaikan permasalahan tersebut dapat memanfaatkan teknologi komputasi di bidang pertanian dengan adanya sistem identifikasi penyakit pada daun cabai berdasarkan warna daun agar dapat memudahkan petani dalam mengetahui penyakit pada daun cabai khususnya penyakit fitoftora, antraknosa dan cercospora, dengan menggunakan algoritma klasifikasi Learning Vector Quantization menghitung jarak antara bobot awal dengan data pelatihan, sehingga di akhir epoch akan menemukan bobot akhir pada setiap kelas. Tahap awal mengaplikasikan metode ini dengan mengumpulkan masing-masing 10 data latih dan data uji masing-masing kategori dengan total data latih 30 data gambar. Data dikumpulkan berupa citra gambar digital yang diolah menggunakan sistem dengan melakukan resize, transfomasi RGB menjadi HSV yang selanjutnya ke proses deteksi tepi canny dengan tujuan mendapatkan pola dari gambar daun cabai. Hasil dari pengujian algoritma Learning Vector Quantization menggunakan confusion matrix mendapatkan hasil akurasi sebesar 80%, nilai precision sebesar 80%, nilai recall sebesar 82%, dan nilai f-1 score sebesar 81%.
Red chili plants are a group of vegetables that are in demand by the people of Indonesia. The problem that often occurs is that there are organisms that interfere with chili plants so that it can make chili production decrease. There are chili plant diseases that are difficult to recognize by farmers using their eyes and without using tools. Chili plant diseases have certain characteristics. In solving these problems, they can utilize computational technology in agriculture with a system of identifying diseases on chili leaves based on leaf color in order to make it easier for farmers to find diseases on chili leaves, especially phytophthora, anthracnose and cercospora diseases, using the Learning Vector Quantization classification algorithm to calculate the distance between initial weights with training data, so that at the end of the epoch you will find the final weight for each class. The initial stage of applying this method is to collect 10 training data and test data for each category with a total of 30 training images. Data were collected in the form of digital images that were processed using the system by resizing, transforming RGB to HSV which then went to the canny edge detection process with the aim of getting the pattern from the chili leaf image. The results of testing the Learning Vector Quantization algorithm using confusion matrix get an accuracy of 80%, a precision value of 80%, a recall value of 82%, and an f-1 score of 81%.
Item Type: | Thesis (Diploma) |
---|---|
Uncontrolled Keywords: | Penyakit daun, Learning Vector Quantization, Confusion Matrix leaf disease, Learning Vector Quantization, Confusion Matrix |
Subjects: | Skripsi Bidang Keilmuan > Teknik Informatika |
Divisions: | Fakultas Telematika Energi > S1 Teknik Informatika |
Depositing User: | Nurul Hidayati |
Date Deposited: | 17 Sep 2025 01:39 |
Last Modified: | 17 Sep 2025 01:39 |
URI: | https://repository.itpln.ac.id/id/eprint/993 |