Ramadhani, Mochammad Rizky (2026) Implementasi algoritma Convolutional Neural Network pada klasifikasi tanaman endemik Pulau Jawa berdasarkan citra daun. Sarjana thesis, UIN Sunan Gunung Djati Bandung.
Full text not available from this repository. (Request a copy)Abstract
INDONESIA: Penelitian ini membahas penerapan algoritma Convolutional Neural Network (CNN) dengan arsitektur InceptionV3 untuk mengklasifikasikan tanaman endemik Pulau Jawa berdasarkan citra daun. Metodologi penelitian menggunakan kerangka kerja CRISP-DM yang mencakup tahapan pemahaman masalah, pengolahan data, pemodelan, dan evaluasi. Dataset yang digunakan terdiri dari 1.000 citra daun dari 10 spesies tanaman yang mewakili flora khas Pulau Jawa, bersumber dari repositori iNaturalist dengan kualitas Research Grade, dan dibagi menjadi data latih, validasi, serta data uji. Model dilatih menggunakan pendekatan transfer learning dengan proses pra-pemrosesan dan augmentasi data untuk meningkatkan kinerja klasifikasi. Hasil pengujian menunjukkan bahwa model mencapai akurasi 89%, dengan nilai precision 90,19%, recall 89%, dan F1-score 89,07%, sehingga dapat disimpulkan bahwa CNN InceptionV3 mampu mengklasifikasikan citra daun tanaman endemik Pulau Jawa secara efektif dan berpotensi digunakan sebagai alat bantu identifikasi awal. ENGLISH: This study implements a Convolutional Neural Network (CNN) using the InceptionV3 architecture to classify endemic plants of Java Island based on leaf images. The research methodology follows the CRISP-DM framework, including problem understanding, data processing, modeling, and evaluation stages. The dataset consists of 1,000 leaf images from 10 plant species representing characteristic flora of Java Island, obtained from the iNaturalist repository with Research Grade quality and divided into training, validation, and testing data. The model was trained using a transfer learning approach with image preprocessing and data augmentation to improve classification performance. Experimental results show that the proposed model achieved an accuracy of 89%, with precision of 90.19%, recall of 89%, and F1-score of 89.07%, indicating that the CNN InceptionV3 model is effective for classifying leaf images of endemic plants and has potential as a preliminary identification tool.
| Item Type: | Thesis (Sarjana) |
|---|---|
| Subjects: | Data Processing, Computer Science > Computer and Human Data Processing, Computer Science > Systems Analysis and Computer Design |
| Divisions: | Fakultas Sains dan Teknologi > Program Studi Teknik Informatika |
| Depositing User: | Mochammad Rizky Ramadhani |
| Date Deposited: | 16 Mar 2026 02:32 |
| Last Modified: | 16 Mar 2026 02:32 |
| URI: | https://digilib.uinsgd.ac.id/id/eprint/128817 |
Actions (login required)
 |
View Item |
