EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC 4.0

COMPARATIVE EVALUATION OF YOLOV5–YOLOV11 MODELS FOR DETECTING NUTRIENT DEFICIENCY IN CHILI SEEDLINGS

Rangga Pebrianto Rangga, Agus Buono, Heru Sukoco, Aziz Kustiyo, Muhamad Syukur

JITK (Jurnal Ilmu Pengetahuan dan Teknologi Komputer) · 26.05.2026 · 11 (4) · 1238-1252

DOI: 10.33480/jitk.v11i4.8263 ↗

Kurz gefasst

Fachartikel in der Zeitschrift JITK (Jurnal Ilmu Pengetahuan dan Teknologi Komputer), online erschienen am 26.05.2026. Verfasst von Rangga Pebrianto Rangga und 4 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Nutrient deficiencies during the seedling stage of chili plants can reduce crop productivity, while conventional identification methods remain subjective and costly. This study compares YOLOv5 to YOLOv11 object detection models for detecting nutrient deficiency symptoms in Bonita chili seedling leaves, including complete nutrition, nitrogen deficiency, phosphorus deficiency, potassium deficiency, and NPK deficiency. The final dataset comprised 4,173 images derived from 1,739 original annotated leaf images through controlled dataset preparation, including split-before-augmentation, laboratory validation of nutrient conditions, and expert-reviewed labeling. All YOLO models were trained and evaluated using the same dataset partition and comparable experimental settings. Performance was assessed using mAP@0.5, computational complexity (FLOPs), inference speed, and model size. The results show that all evaluated models achieved high detection performance, with differences mainly appearing in computational efficiency and the balance between accuracy and speed. YOLOv10s and YOLOv11s obtained the highest mAP@0.5 in this experiment, whereas YOLOv8s showed a competitive balance between accuracy, inference speed, and model compactness. These findings indicate that recent YOLO developments are promising for fine-grained nutrient deficiency detection in computer vision–based precision agriculture.

Zitieren

APA

Rangga, R. P., Buono, A., Sukoco, H., Kustiyo, A., & Syukur, M. (2026). COMPARATIVE EVALUATION OF YOLOV5–YOLOV11 MODELS FOR DETECTING NUTRIENT DEFICIENCY IN CHILI SEEDLINGS. JITK (Jurnal Ilmu Pengetahuan dan Teknologi Komputer), 11(4), 1238-1252. https://doi.org/10.33480/jitk.v11i4.8263

Vancouver

Rangga RP, Buono A, Sukoco H, Kustiyo A, Syukur M. COMPARATIVE EVALUATION OF YOLOV5–YOLOV11 MODELS FOR DETECTING NUTRIENT DEFICIENCY IN CHILI SEEDLINGS. JITK (Jurnal Ilmu Pengetahuan dan Teknologi Komputer). 2026;11(4):1238-1252. doi:10.33480/jitk.v11i4.8263

BibTeX
@article{Rangga2026,
  author = {Rangga, Rangga Pebrianto and Buono, Agus and Sukoco, Heru and Kustiyo, Aziz and Syukur, Muhamad},
  title = {COMPARATIVE EVALUATION OF YOLOV5–YOLOV11 MODELS FOR DETECTING NUTRIENT DEFICIENCY IN CHILI SEEDLINGS},
  journal = {JITK (Jurnal Ilmu Pengetahuan dan Teknologi Komputer)},
  year = {2026},
  volume = {11},
  number = {4},
  pages = {1238--1252},
  doi = {10.33480/jitk.v11i4.8263}
}
Kostenlos starten →