Innovations in Science and Technologies Ҷилди 1 № 1 (2024) · Саҳифаҳои 60-70

OBJECT DETECTION AND DISTANCE MEASUREMENT

Baymatova, M.X., Nuratdinova, K., Raxmanov, M.

DOI: 10.5281/zenodo.10824703 Дар манбаъ хондан PDF Иқтибос гирифтан

Аннотатсия

Object detection and distance measurement are fundamental tasks in computer vision, with applications ranging from autonomous vehicles to surveillance systems. This paper provides an overview of the various techniques andtechnologies used for object detection and distance measurement, including their principles, advantages, and limitations. We discuss the importance of combining these two capabilities to extract valuable information for real-world applications. We used Yolo4 tiny for project. YOLOv4-tiny is the compressed version of YOLOv4. The YOLOv4-tiny model achieves 22.0% AP (42.0% AP50) at a speed of 443 FPS on RTX 2080Ti, while by using TensorRT, batch size = 4 and FP16- precision the YOLOv4-tiny achieves 1774 FPS. Moreover, in order to create project, we utilized range of methods such as autoencoder, CNN and DNN.

YOLOv4-tiny; One-stage methods; Two-stage methods; autoencoder, CNN, DNN.

Манбаи метамаълумот: бойгонии OAI-PMH-и маҷалла · Sindex матни пурраро нигоҳ намедорад, ба манбаъ пайванд медиҳад.