Length Estimation of Moving Objects With Ann and Gripping With Robotic Arm
Yükleniyor...
Dosyalar
Tarih
2023 Aralık
Yazarlar
Dergi Başlığı
Dergi ISSN
Cilt Başlığı
Yayıncı
Selçuk Üniversitesi
Erişim Hakkı
info:eu-repo/semantics/openAccess
Özet
It is possible to obtain general information about objects with image processing. However, while
measuring the size of objects, especially with 2D cameras, the calibrated systems have been
worked with restrictions such as fixed length and distance. However, without depth information
and for objects of arbitrary positions and lengths, calculating their dimensions is a rather difficult
task. In this study, an ANN-based application was carried out to calculate the amount of movement
and the length of the object by viewing the moving objects from the side and top with two cameras.
Objects moving on the conveyor belt are detected by deep learning-based YOLO. The motion
amount of the object was calculated in the second image with the template created on the detected
objects. An ANN is trained with the amount of movement and position information measured by
two cameras. At the end of the training, the network estimates the lengths of the objects with small
errors. The speed of the objects was calculated according to the calculated length and the targets
were grasped with a robot arm.
Açıklama
Anahtar Kelimeler
Artificial Neural Network, Grasping, Moving object, Object dimension
Kaynak
Selcuk University Journal of Engineering Sciences
WoS Q Değeri
Scopus Q Değeri
Cilt
22
Sayı
3
Künye
Uçar, K., Koçer, H. E., (2023). Length Estimation of Moving Objects With Ann and Gripping With Robotic Arm. Selcuk University Journal of Engineering Sciences, 22(3), 131-136.
