Object Recognition System for the Spinbotics Robotic Arm

Fulltext (.pdf 1,24 MB)

https://doi.org/10.52651/sam.a.2024.1.39-44

Autor: Patrik Štefka, Peter Pásztó, Marian Kľúčik, Martin Smoľák, Matej Vargovčík, Jakub Lenner

In: Science & Military

ISSN: 2453-7632

Ročník: 19

Číslo: 1

Strany: 39 - 44

Rok vydania: 2024

Vydavateľ: Akadémia ozbrojených síl generála Milana Rastislava Štefánika, Liptovský Mikuláš

Abstrakt: This study focuses on the development of a visual system designed to facilitate object detection for the Spinbotics robotic arm in spatial environments. The primary objective is to enable accurate detection and classification of diverse objects, enhancing the arm's capability to grasp and manipulate items effectively. The system employs the YOLOv7 deep neural network, fine-tuned using transfer learning on a local computing infrastructure. Compared to traditional methods like R-CNN and SSD, YOLOv7 offers superior real-time processing capabilities and efficiency, making it well-suited for dynamic environments. Through extensive training and testing, the system demonstrates robust performance in detecting objects across
varied scenes and identifying optimal grasp points. This research underscores the effectiveness of integrating advanced computer vision techniques to enhance the operational efficiency and versatility of robotic manipulators in real-world applications.

Kľúčové slová: Robot arm; Visual system; YOLO; Object detection.

Citácia:

ŠTEFKA, Patrik; PÁSZTÓ, Peter; KĽÚČIK, Marian; SMOĽÁK, Martin; VARGOVČÍK, Matej et al. Object Recognition System for the Spinbotics Robotic Arm. In: Science & Military [online]. Liptovský Mikuláš: Akadémia ozbrojených síl generála Milana Rastislava Štefánika, 2024, 19 (1), s. 39-44. ISSN: 1336-8885 (print). ISSN: 2453-7632 (online). Dostupné na internete: https://doi.org/10.52651/sam.a.2024.1.39-44