Computational Science Hub (CSH)

Permanent URI for this collectionhttps://hohpublica.uni-hohenheim.de/handle/123456789/16924

Browse

Browsing Computational Science Hub (CSH) by Subject "Computer vision"

Type the first few letters and click on the Browse button
Now showing 1 - 3 of 3
  • Thumbnail Image
    Publication
    Assessing the capability of YOLO- and transformer-based object detectors for real-time weed detection
    (2025) Allmendinger, Alicia; Saltık, Ahmet Oğuz; Peteinatos, Gerassimos G.; Stein, Anthony; Gerhards, Roland
  • Thumbnail Image
    Publication
    Assessing the capability of YOLO- and transformer-based object detectors for real-time weed detection
    (2025) Allmendinger, Alicia; Saltık, Ahmet Oğuz; Peteinatos, Gerassimos G.; Stein, Anthony; Gerhards, Roland
    Spot spraying represents an efficient and sustainable method for reducing herbicide use in agriculture. Reliable differentiation between crops and weeds, including species-level classification, is essential for real-time application. This study compares state-of-the-art object detection models-YOLOv8, YOLOv9, YOLOv10, and RT-DETR-using 5611 images from 16 plant species. Two datasets were created, dataset 1 with training all 16 species individually and dataset 2 with grouping weeds into monocotyledonous weeds, dicotyledonous weeds, and three chosen crops. Results indicate that all models perform similarly, but YOLOv9s and YOLOv9e, exhibit strong recall (66.58 % and 72.36 %) and mAP50 (73.52 % and 79.86 %), and mAP50-95 (43.82 % and 47.00 %) in dataset 2. RT-DETR-l, excels in precision reaching 82.44 % (dataset 1) and 81.46 % (dataset 2) making it ideal for minimizing false positives. In dataset 2, YOLOv9c attains a precision of 84.76% for dicots and 78.22% recall for Zea mays L.. Inference times highlight smaller YOLO models (YOLOv8n, YOLOv9t, and YOLOv10n) as the fastest, reaching 7.64 ms (dataset 1) on an NVIDIA GeForce RTX 4090 GPU, with CPU inference times increasing significantly. These findings emphasize the trade-off between model size, accuracy, and hardware suitability for real-time agricultural applications.
  • Thumbnail Image
    Publication
    Poultry perfection : comparison of computer vision models to detect and classify poultry products in a production setting
    (2025) Einsiedel, Daniel; Vita, Marco; Jox, Dana; Dunnewind, Bertus; Meulendijks, Johan; Krupitzer, Christian
    This study explores the use of computer vision, specifically object detection, for quality control in ready-to-eat meat products. We focused on a single process step, labeling products as “good” or “imperfect”. An “imperfect product” constitutes a product that deviates from the norm regarding shape, size, or color (having a hole, missing edges, dark particles, etc.). Imperfect does not mean the product is inedible or a risk to food safety, but it affects the overall product quality. Various object detectors, such as YOLO, including YOLO12, were compared using the mAP50-95 metric. Most models achieved mAP scores over 0.9, with YOLO12 reaching a peak score of 0.9359. The precision and recall curves indicated that the model learned the “imperfect product” class better, most likely due to its higher representation. This underscores the importance of a balanced dataset, which is challenging to achieve in real-world settings. The confusion matrix revealed false positives, suggesting that increasing dataset volume or hyperparameter tuning could help. However, increasing the dataset volume is usually the more difficult path since data acquisition and especially labeling are by far the most time-consuming steps of the whole process. Overall, current models can be applied to quality control tasks with some margin of error. Our experiments show that high-quality, consistently labeled datasets are potentially more important than the choice of the model for achieving good results. The applied hyperparameter tuning on the YOLO12 model did not outperform the default model in this case. Future work could involve training models on a multi-class dataset with hyperparameter optimization. A multi-class dataset could contain more specific classes than just “good” and “imperfect,” making trained models capable of actually predicting specific quality deviations.