Scholarlink

<p>The purpose of smart parking systems is to optimize the use of parking spaces in urban areas by combining sensors, cameras, and software. The systems detect parking spots automatically and notify users in real-time, instead of requiring manual searches for parking.They contribute to the reduction of traffic congestion, fuel consumption, and environmental pollution by reducing time spent searching for parking. This paper proposes a vision-based smart parking system that utilizes deep learning techniques for real-time detection and monitoring of parking space occupancy. The system employs YOLOv7 for object detection, integrated with a Flask-based web interface and OpenCV for image preprocessing. The SQLite database makes it possible for a responsive web application to deliver timely, accurate information with low latency to users. The experimental results presented in this paper indicate that the system achieved an overall average detection accuracy of 94.23% across diverse environmental conditions, which reflects the project’s scalability and dependability. Furthermore, our approach offers a low-cost, easy to implement solution with minimal dependence on hardware-based infrastructure that can be applied in a wide range of land use in urban contexts. Furthermore, the project aligns with securing critical infrastructure in next generation networks, promoting urban mobility, enhancing intelligent transportation systems, and ensuring resilience of smart urban services that are resource-efficient and accessible. This AI-enhanced framework presents an opportunity to protect and optimize critical urban infrastructure using AI-based technology, while also using machine learning in its real-time decisions.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Deep learning enabled real time parking monitoring using YOLOv7 for intelligent and secure critical infrastructure

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年12月16日

查看全文
  • Sonia Verma,
  • Abhilasha Singh,
  • M. P. Sunil,
  • Sardar M. N. Islam,
  • D. Satish Kumar,
  • M. R. Ebenezar Jebarani

摘要

The purpose of smart parking systems is to optimize the use of parking spaces in urban areas by combining sensors, cameras, and software. The systems detect parking spots automatically and notify users in real-time, instead of requiring manual searches for parking.They contribute to the reduction of traffic congestion, fuel consumption, and environmental pollution by reducing time spent searching for parking. This paper proposes a vision-based smart parking system that utilizes deep learning techniques for real-time detection and monitoring of parking space occupancy. The system employs YOLOv7 for object detection, integrated with a Flask-based web interface and OpenCV for image preprocessing. The SQLite database makes it possible for a responsive web application to deliver timely, accurate information with low latency to users. The experimental results presented in this paper indicate that the system achieved an overall average detection accuracy of 94.23% across diverse environmental conditions, which reflects the project’s scalability and dependability. Furthermore, our approach offers a low-cost, easy to implement solution with minimal dependence on hardware-based infrastructure that can be applied in a wide range of land use in urban contexts. Furthermore, the project aligns with securing critical infrastructure in next generation networks, promoting urban mobility, enhancing intelligent transportation systems, and ensuring resilience of smart urban services that are resource-efficient and accessible. This AI-enhanced framework presents an opportunity to protect and optimize critical urban infrastructure using AI-based technology, while also using machine learning in its real-time decisions.