A survey study in Object Detection: A Comprehensive Analysis of Traditional and State-of-the-Art Approaches

Abstract

Object detection is an essential field within computer vision, focusing on identifying objects' presence and category within image or video data. The significance of this issue is paramount in numerous domains that directly impact people's lives, including autonomous driving, healthcare systems, and security monitoring. In contrast to traditional methodologies employed for object detection, deep learning-based algorithms have demonstrated substantial progress in computational efficiency and precision in recent years. This study aims to provide a comprehensive review of object detection by methodically employing deep learning to facilitate a comprehensive and in-depth comprehension of the fundamental principles in this field. The discussion has encompassed various subjects, such as the obstacles and complexities associated with object detection and the traditional and deep learning detectors. The detection of objects within images and videos, the real-time detection of objects, detection of 3D objects, commonly used datasets, and the metrics employed for evaluating object detection performance. This study will likely yield scientific benefits for academics working in the field of object detection and deep learning.