Research & Innovation
Intelligent Edelweiss Health Detection System Based on Digital Images
Fundamental Basic Research Scheme · Universitas Nusa Putra
Edelweiss (Anaphalis javanica) is an endemic mountain flora of Indonesia, threatened by climate change and tourism activity. Its health monitoring has been done manually, making it subjective and inefficient. This research develops an intelligent image-based system combining YOLOv11 for flower object detection and a Multi-Layer Perceptron (MLP) for health condition classification, in support of endemic plant conservation.
Research Highlights
Two-Stage Method
YOLOv11 detects & localizes flowers, then MLP classifies their health level.
Field Dataset
3,000 Edelweiss flower images collected directly from natural habitat at Mount Gede Pangrango and Mount Lawu.
Research Team
The researchers behind the development of this system.
Anggun Fergina, M.Kom
Ketua Peneliti
Dosen — Universitas Nusa Putra
Aulia Kusuma, S.Tr.Tra., M.MT.
Anggota Peneliti
Dosen — Politeknik Negeri Batam
M. Ikhsan Thohir, S.Kom., M.Kom
Anggota Peneliti
Dosen — Universitas Nusa Putra
Siti Marni
Anggota Peneliti
Mahasiswa — Universitas Nusa Putra
Puput Handayani
Anggota Peneliti
Mahasiswa — Universitas Nusa Putra
Ervin Agustian Gunawan
Anggota Peneliti
Mahasiswa — Universitas Nusa Putra
Data Collection Locations
The photos used to train this system were collected directly from two mountains in Indonesia that are natural habitats of Java Edelweiss. You can view the locations on the map.
Finding Flowers in the Photo
The system first locates where the Edelweiss flowers are in the image, then marks them with boxes. It works like an eye that recognizes each individual bloom.
Determining Health Condition
Once flowers are found, the system inspects each one and decides whether it is Blooming, in Full Bloom, or still in the Seedling phase.