In this research we focus on using deep learning for the training of real time detection of defected Nails and Nuts on a high speed production line using You Only Look Once (YOLO) algorithm for real time object detection and trying to increase the precision of detection and decrease the problems facing real time object detection models like Object occlusion, different orientation for objects, lighting conditions, undetermined moving objects and noise. A series of experiments have been done to achieve high prediction accuracy, the experimental results made on our costumed pascal visual object classes (VOC) dataset demonstrated that the mean Average Precision (mAP) could reach 85%. The proposed model showed very good prediction accuracy on the test dataset.
CITATION STYLE
Teama, T., Ma, H., Maher, A., & Kassab, M. A. (2019). Real Time Object Detection Based on Deep Neural Network. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 11743 LNAI, pp. 493–504). Springer Verlag. https://doi.org/10.1007/978-3-030-27538-9_42
Mendeley helps you to discover research relevant for your work.