Conteo y reconocimiento de actividad de personas en un recinto mediante image processing

With the Covid-19 pandemic overcrowdedness in closed spaces becomes a main issue to manage, specially in places with high transit of people like universities, malls, public offices, etc. With the help of image processing it is going to be developed a algorithm that will provide real time data on occupation of specific places, also it is going to provide a activity monitor tagging people in the same room and what they are doing. This data will be exposed through standard interfaces to be consumed by information systems. Finally the objective is deploying this on embedded systems. Also it is going to be developed a dashborad to be able to visualise the data that is been processed in realtime. The extend of this project is mainly scoped under the university use case but it can be extended to other public places like malls, public offices, hospitals and more. Using this technics opens possibilities to make better decisions on space optimisation and give realtime data for users that share this types of spaces.

• Autores: Arellano Ramírez, Juan Felipe
• Fecha de publicación: 2021
• Institución: Universidad del Norte
• Repositorio: Repositorio Uninorte
• Idioma: spa
• OAI Identifier: oai:manglar.uninorte.edu.co:10584/10295
• Acceso en línea: http://hdl.handle.net/10584/10295
• Palabra clave: image processing; people counting; HAR; Human activity recognition; embedded; procesamiento de imagen; conteo de personas; Reconocimiento de actividad humana; dispositivos integrados
• License: Universidad del Norte