Article Per Year (5 Year)
p-Index From 2021 - 2026
2.697
P-Index
This Author published in this journals
All Journal Network Engineering Research Operation [NERO] JITK (Jurnal Ilmu Pengetahuan dan Komputer) Jurnal Sisfokom (Sistem Informasi dan Komputer) Jurnal TIPS : Jurnal Teknologi Informasi dan Komputer Politeknik Sekayu JSAI (Journal Scientific and Applied Informatics) Informanika Ngabdimas : Pengabdian Pada Masyarakat JPMTT (Jurnal Pengabdian Masyarakat Teknologi Terbarukan) JCOSIS (Journal Computer Science and Information Systems)
Mariana Purba
Unknown Affiliation
Civil Engineering, Building, Construction & Architecture Computer Science & IT Control & Systems Engineering Decision Sciences, Operations Research & Management Economics, Econometrics & Finance Education Languange, Linguistic, Communication & Media Other

Published : 16 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 4 Documents
Search
Journal : jcosis journal computer science and information systems

 1 
Klasifikasi Jenis Daun Tanaman Tropis Menggunakan Model ResNet50 Berbasis Transfer Learning dengan Dataset Tropical Plant Leaf Mariana Purba
JCOSIS (Journal Computer Science and Information Systems) Vol. 2 No. 2 (2025): Oktober
Publisher : Institute for Research and Community Service

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.61567/jcosis.v2i2.255

Abstract

Identifikasi otomatis jenis tanaman melalui klasifikasi citra daun memiliki peran penting dalam mendukung pertanian cerdas dan konservasi keanekaragaman hayati. Penelitian ini mengusulkan penerapan model ResNet50 berbasis transfer learning untuk klasifikasi jenis daun tanaman tropis dengan menggunakan Tropical Plant Leaf Dataset. Dataset ini terdiri atas ribuan citra daun dari berbagai spesies tanaman tropis yang telah melalui proses preprocessing seperti resizing, normalisasi, dan augmentasi data untuk meningkatkan kemampuan generalisasi model. Model ResNet50 yang digunakan merupakan hasil fine-tuning dari bobot pra-latih ImageNet, dilatih selama 50 epoch menggunakan optimizer Adam dan fungsi kerugian categorical cross-entropy. Hasil eksperimen menunjukkan bahwa model ResNet50 mencapai akurasi sebesar 99,45%, presisi 99,40%, recall 99,35%, dan F1-score 99,32%, serta mengungguli model CNN konvensional dan VGG16. Visualisasi Grad-CAM memperlihatkan bahwa model mampu mengenali pola venasi dan tekstur daun yang relevan secara biologis dalam proses klasifikasi. Temuan ini membuktikan bahwa ResNet50 efektif digunakan untuk klasifikasi citra daun tanaman tropis dan dapat menjadi dasar pengembangan sistem identifikasi tanaman otomatis di lingkungan pertanian tropis. Penelitian selanjutnya disarankan untuk mengintegrasikan metode ensemble learning atau attention mechanism guna meningkatkan ketahanan model terhadap variasi pencahayaan dan latar belakang citra.
Implementasi Metode Color Blob Detection Pada Objek Daun Sawi Wahyu Adianto; Dwika Putra, Erwin; Handrie Noprison; Vina Ayumi; Marissa utami; Mariana Purba
JCOSIS (Journal Computer Science and Information Systems) Vol. 1 No. 1 (2024): Mei
Publisher : Institute for Research and Community Service

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.61567/jcosis.v1i1.176

Abstract

Tujuan : Sistem aplikasi Mendeteksi Objek Daun Sawi dengan menerapakan Metode Color Blob Detection menggunakan Bahasa pemrograman matlab. Mekanisme pembuatan aplikasi dimulai dari pembuatan rancangan aplikasi, GUI (Grafik User Interface) sampai dengan pembuatan coding. Metode/Design/Pendekatan: model deteksi Objek Daun Sawi dengan menerapakan Metode Color Blob Detection menggunakan Bahasa pemrograman matlab Hasil/Temuan: Hasil pengujian yang dilakukan dengan memiliki tingkat keakuratan paling tinggi yaitu dengan jarak pengambilan sampel objek dengan jarak 50 cm, dan tingkat keakuratan paling rendah dengan jarak pengambilan sampel objek dengan jarak 20 cm. Kebaharuan/Originalitas/Nilai: Tingkat akurasi deteksi daun sawi maka dapat disimpulkan berhasil dengan tingkat keberhasil akurasi 67.7% Keywords: Color Blob Detection, Image Processing, Matlab
Penerapan Algoritma Viola Jones Untuk Deteksi Mata Pada Pengolahan Citra Digital Putra, Erwin Dwika; Marissa Utami; Mariana Purba
JCOSIS (Journal Computer Science and Information Systems) Vol. 1 No. 2 (2024): Oktober
Publisher : Institute for Research and Community Service

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.61567/jcosis.v1i2.208

Abstract

Purpose: This research aims to test the performance of eye detection algorithms in human images using specific methods, such as the Viola-Jones algorithm. Eye detection is an important component in various applications, ranging from biometrics to surveillance systems. Methods/Study design/approach: In this study, testing was conducted on 10 images with simple variations to assess the algorithm's accuracy, precision, and recall. Result/Findings: The results showed that the algorithm achieved an accuracy rate of 80% with a precision and recall of 88.9%, which indicates a fairly reliable performance for standard conditions. However, detection errors such as false positives and negatives were found, suggesting the potential for improvement, especially in images with varying illumination or orientation. Novelty/Originality/Value: This algorithm is suitable for real-time eye detection applications with stable lighting and orientation conditions. Still, it is recommended to incorporate additional methods to improve accuracy in more complex conditions.
IMPLEMENTASI USER ACCEPTANCE TESTING (UAT) PADA RANCANG BANGUN SISTEM INFORMASI PENERIMAAN PESERTA DIDIK BARU (PPDB) Dwi Diana Sari; Mariana Purba; Nia Umilizah
JCOSIS (Journal Computer Science and Information Systems) Vol. 1 No. 2 (2024): Oktober
Publisher : Institute for Research and Community Service

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.61567/jcosis.v1i2.209

Abstract

Purpose: The purpose of this test is to ensure that the main features of the system meet the needs and specifications that have been set. From a total of 9 test scenarios, 7 scenarios were successfully passed, while 2 scenarios failed, namely in the Document Upload feature (KK, Ijazah) and Print Proof of Registration. Methods/Study design/approach: The results of User Acceptance Testing (UAT) - Alpha Testing for the New Student Admission Information System (PPDB) conducted by the internal QA team at the research object, namely SMK Negeri 3 Kayu Agung. Result/Findings: This failure indicates the need for further development, as both features are crucial elements in the registration process. Based on the test results, future research is recommended to address these issues immediately so that the system can operate properly. The final decision concludes that the system is not ready for Beta Testing until all issues are resolved. Novelty/Originality/Value: This research with UAT evaluation aims to provide insights for future research and development regarding the readiness status of the system as well as the corrective measures needed to improve the user experience.
Co-Authors Ade Sukma Wati Aulia Aryani Ayumi, Vina Bakhtiar Bakhtiar .K Bakhtiar K Bakhtiar, Bakhtiar. Bela Okta Riyanto Cahyani, Eni Dewi Oktafian Dian Nugraha Dwi Diana Sari Dwi Fitri Brianna Eldhiyano Pratama Hadiguna Setiawan Handrie Noprison Junaidi Junaidi Lemi Iryani Lemi Iryani Meslady .K Nia Umilizah Novita Nurjannah Putri Nur Ani Oktafiani, Dewi Paisal Pipin Octavia Putra, Erwin Dwika Rizki Adi Saputra Syamsiar, Syamsiar Wachyu Hari Haji Wahyu Adianto