Daylight Investigations

Instance Segmentation and Daylight


Valeria Belén Cerpa Salas


Roberto Flores Gutiérrez


Universidad Católica de Santa María



Instance Segmentation and Daylight

Project Description

In many Latin American cities, as well as other cities in different regions of the world, there is the problem of insecurity. This risk mainly affects women, who travel narrow streets and alleys for obligatory reasons during the morning or afternoon. Reasons for this inadequate lighting is due to the presence of tall buildings that do not respect municipal regulations, continuous walls, and inadequate urban planning. In response to this, a question arises: How do we know if a public space has adequate daylight? With the support of Computer Science and specifically of the “Instance Segmentation” process, we will be able to analyze through photographs of alleys or narrow streets. These images could be taken at different times of the day, and it will show the incidence of sunlight on surfaces. Although Computer Science seems a completely different area in comparison with