Analyzing Real-Time Object Detection with YOLO Algorithm in Automotive Applications: A Review

Abstract

Identifying objects in real-time is a technology that is developing rapidly and has a huge potential for expansion in many technical fields. Currently, systems that use image processing to detect objects are based on the information from a single frame. A video camera positioned in the analyzed area captures the image, monitoring in detail the changes that occur between frames. The You Only Look Once (YOLO) algorithm is a model for detecting objects in images, that is currently known for the accuracy of the data obtained and the fast-working speed. This study proposes a comprehensive literature review of YOLO research, as well as a bibliometric analysis to map the trends in the automotive field from 2020 to 2024. Object detection applications using YOLO were categorized into three primary domains: road traffic, autonomous vehicle development, and industrial settings. A detailed analysis was conducted for each domain, providing quantitative insights into existing implementations. Among the various YOLO architectures evaluated (v2 - v8, H, X, R, C), YOLO v8 demonstrated superior performance with a mean Average Precision (mAP) of 0.99.