An Application for Research of Object Selection and Tracking in Images DOI
В. Власенко, Andrii Dashkevych, Darya Vorontsova

и другие.

Опубликована: Окт. 2, 2023

The work is devoted to the process of research and development an object selection tracking in images. article offers a solution problem crossing bounding frames that arises operation YOLO R-CNN type neural networks by developing method for estimating optimal size search area, which will allow finding appropriate its frame, offer generalized approach visualization process, visually represent overlap facilitate object. To confirm effectiveness proposed method, there conduct experiments on data set.

Язык: Английский

Aerial Imagery to Support Participatory Landscape Decision-Making DOI
Yilin Huang, Eva Lieberherr,

Khammeun Nandee

и другие.

Опубликована: Янв. 1, 2024

Язык: Английский

Процитировано

0

Drone Insights: Unveiling Beach Usage through AI-Powered People Counting DOI Creative Commons
César Herrera, Rod M. Connolly, J. B. Rasmussen

и другие.

Drones, Год журнала: 2024, Номер 8(10), С. 579 - 579

Опубликована: Окт. 13, 2024

Ocean beaches are a major recreational attraction in many coastal cities, requiring accurate visitor counts for infrastructure planning and value estimation. We developed novel method to assess beach usage on the Gold Coast, Australia, using 507 drone surveys across 24 beaches. The covered 30 km of coastline, accounting different seasons, times day, environmental conditions. Two AI models were employed: one counting people land water (91–95% accuracy), another identifying types (85–92% accuracy). Using data, we estimated annual at 34 million 2022/23, with 55% 45% water—approximately double most recent estimate from lifeguard counts, which spatially limited prone human error. When applying similar restrictions as surveys, data 15 visits, aligning closely (within 9%). Temporal (time day week, season) spatial (beach location) factors strongest predictors usage, additional patterns explained by weather variables. Our method, combining drones AI, enhances coverage, accuracy, granularity monitoring, offering scalable, cost-effective solution long-term assessment.

Язык: Английский

Процитировано

0

Life of Drone Visuals: Norms, Ethics, and Effects DOI
Elisa Serafinelli

Опубликована: Янв. 1, 2024

Язык: Английский

Процитировано

0

El cambio de cobertura y uso del terreno desde el enfoque de la metodología mixta: una revisión de la literatura DOI Creative Commons

Lucero Pimienta Ramírez,

Erna López Granados

LATAM Revista Latinoamericana de Ciencias Sociales y Humanidades, Год журнала: 2023, Номер 4(2)

Опубликована: Авг. 2, 2023

Este artículo ofrece un análisis de la literatura internacional producido entre el año 2012 y 2022 sobre abordaje del Cambio Cobertura Uso Terreno (CCUT), a través una metodología mixta. El fenómeno involucra múltiples interacciones con los factores físicos, sociales, económicos, políticos culturales. Por esta razón, se requiere comprender manera integral las causas, efectos procesos que inducen estos cambios nivel local regional. La mixta perspectiva combina datos cuantitativos (teledetección) cualitativos (percepción social). objetivo investigación fue revisión estudios científicos utilizan aplicada al CCUT, para conocer estado actual conocimiento enfoques teóricos, metodológicos, tendencias en este campo investigación. Se encontró limitada revistas acceso abierto aborden problemática enfoque, diversidad término “metodología mixta”, no existe información explícita método artículos revisados. identificaron herramientas técnicas más empleadas cuantitativa cualitativa, estrategias, alcances limitaciones han reportado diversos estudios. destaca capacidad enfoque metodológico obtener comprensión problemática, debido proporciona completa causas conducen CCUT. También identificó limitación importante como establecer vínculos previos confianza actores locales puedan colaborar diseño ser parte fundamental

Процитировано

0

An Application for Research of Object Selection and Tracking in Images DOI
В. Власенко, Andrii Dashkevych, Darya Vorontsova

и другие.

Опубликована: Окт. 2, 2023

The work is devoted to the process of research and development an object selection tracking in images. article offers a solution problem crossing bounding frames that arises operation YOLO R-CNN type neural networks by developing method for estimating optimal size search area, which will allow finding appropriate its frame, offer generalized approach visualization process, visually represent overlap facilitate object. To confirm effectiveness proposed method, there conduct experiments on data set.

Язык: Английский

Процитировано

0