REAL-TIME DETECTION OF WILD MUSTARD (Sinapis arvensis L.) WITH DEEP LEARNING (YOLO-v3)
Yazarlar (2)
Dr. Öğr. Üyesi Mustafa GÜZEL
Tokat Gaziosmanpaşa Üniversitesi, Türkiye
Tokat Gaziosmanpaşa Üniversitesi, Türkiye
| Makale Türü | Özgün Makale |
| Makale Alt Türü | SSCI, AHCI, SCI, SCI-Exp dergilerinde yayınlanan tam makale |
| Dergi Adı | FRESENIUS ENVIRONMENTAL BULLETIN |
| Dergi ISSN | 1018-4619 |
| Dergi Tarandığı Indeksler | SCI |
| Dergi Grubu | Q4 |
| Makale Dili | İngilizce |
| Basım Tarihi | 01-2021 |
| Cilt No | 30 |
| Sayı | 11 |
| Sayfalar | 12197 / 12203 |
| Makale Linki | https://www.researchgate.net/publication/356065052_Real-time_Detection_of_Wild_Mustard_Sinapis_arvensis_L_With_Deep_Learning_Yolo-v3 |
| Özet |
| In this study, we detected Wild Mustard (Sinapis arvensis L.) plants with deep learning method in real time. The wild mustard (Sinapis arvensis L.) is a parasite plant which caused great losses in wheat farming. The images of wild mustard plants were obtained from the wheat testing area of Gaziosmanpasa University in the spring of 2017-2018 with video recording by drone. The images that used to train the system has taken from the wild mustard video frames were reproduced with data augmentation method. There are totally 8 536 images has created for train and test the deep learning architect. The detection of wild mustard has done using Python 3.7.2 and the YOLO (You Only Look Once) library. We have created two different weights with deep learning and tested on images. The current average losses found %2.32 with 30 k iteration and %1.83 with 100 k iteration. Predicted success rate has varied … |
| Anahtar Kelimeler |
| Deep learning | real time detection | wild mustard | yolov3 |
