Optimizing accuracy and efficiency in real-time people counting with cascaded object detection DOI Creative Commons
M. Raviraja Holla,

D. Suma,

M. Darshan Holla

et al.

International Journal of Information Technology, Journal Year: 2024, Volume and Issue: unknown

Published: Aug. 23, 2024

Abstract Growing concerns about public safety have driven the demand for real-time surveillance, particularly in monitoring systems like people counters. Traditional methods heavily reliant on facial detection face challenges due to complex nature of features. This paper presents an innovative counting system known its robustness, utilizing holistic bodily characteristics improved and tallying. achieves exceptional performance through advanced computer vision techniques, with a flawless accuracy precision rate 100% under ideal conditions. Even challenging visual conditions, it maintains impressive overall 98.42% 97.51%. Comprehensive analyses, including violin plot heatmaps, support this outstanding performance. Additionally, by assessing execution time concerning number cascading stages, we highlight significant advantages our approach. Experimentation TUD-Pedestrian dataset demonstrates 94.2%. Evaluation using UCFCC further proves effectiveness approach handling diverse scenarios, showcasing robustness real-world crowd applications. Compared benchmark approaches, proposed efficiency.

Language: Английский

Optimizing accuracy and efficiency in real-time people counting with cascaded object detection DOI Creative Commons
M. Raviraja Holla,

D. Suma,

M. Darshan Holla

et al.

International Journal of Information Technology, Journal Year: 2024, Volume and Issue: unknown

Published: Aug. 23, 2024

Abstract Growing concerns about public safety have driven the demand for real-time surveillance, particularly in monitoring systems like people counters. Traditional methods heavily reliant on facial detection face challenges due to complex nature of features. This paper presents an innovative counting system known its robustness, utilizing holistic bodily characteristics improved and tallying. achieves exceptional performance through advanced computer vision techniques, with a flawless accuracy precision rate 100% under ideal conditions. Even challenging visual conditions, it maintains impressive overall 98.42% 97.51%. Comprehensive analyses, including violin plot heatmaps, support this outstanding performance. Additionally, by assessing execution time concerning number cascading stages, we highlight significant advantages our approach. Experimentation TUD-Pedestrian dataset demonstrates 94.2%. Evaluation using UCFCC further proves effectiveness approach handling diverse scenarios, showcasing robustness real-world crowd applications. Compared benchmark approaches, proposed efficiency.

Language: Английский

Citations

2