Scholarlink

Crowd counting is the area of computer vision that addresses the problem of detecting and counting the number of people in crowded spaces and is widely used in the areas of application such as public safety and town planning. This work proposes a novel crowd counting approach based on a deep learning model that combines the Mask Region-based Convolution Neural Network (Mask R-CNN) and the Inception ResNet V2 model. The proposed model performs high-precision instance segmentation with the help of feature extraction using Inception ResNet V2. As a result, the model can correctly recognize and count crowded individuals. The paper further outlines the scalability and flexibility of the proposed model, which accommodates different crowd densities and environmental conditions, creating various applications including urban surveillance, event organizing, and public law enforcers. The proposed model detects people in crowd with an accuracy of 98%.

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

Crowd Counting Using Mask R-CNN Inception ResNet V2 Deep Learning Model

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • T. M. Praneeth Naidu,
  • Chandra Sekhar Paidimarry

摘要

Crowd counting is the area of computer vision that addresses the problem of detecting and counting the number of people in crowded spaces and is widely used in the areas of application such as public safety and town planning. This work proposes a novel crowd counting approach based on a deep learning model that combines the Mask Region-based Convolution Neural Network (Mask R-CNN) and the Inception ResNet V2 model. The proposed model performs high-precision instance segmentation with the help of feature extraction using Inception ResNet V2. As a result, the model can correctly recognize and count crowded individuals. The paper further outlines the scalability and flexibility of the proposed model, which accommodates different crowd densities and environmental conditions, creating various applications including urban surveillance, event organizing, and public law enforcers. The proposed model detects people in crowd with an accuracy of 98%.