Activity Funded
Automating City Accessibility Mapping using AI
Automatizar o Mapeamento de Acessibilidades Urbanas com IA
Details
Reference
2022.09218.PTDC
2022.09218.PTDC
Project Start Date
2023-03-01
2023-03-01
Project End Date
2025-02-28
2025-02-28
Scientific Area
Exact sciences
Exact sciences
Funding Program
Concurso de Projetos de I&D em Todos os Domínios Científicos - 2022 - PEX
Concurso de Projetos de I&D em Todos os Domínios Científicos - 2022 - PEX
Abstract
This exploratory project proposes to develop a prototype platform that relies on artificial intelligence and machine learning to automatically detect accessibility features in public space, particularly sidewalks. The platform will rely on OpenStreetMap to collect and identify sidewalks and crossroads in downtown Porto. Then, the Google Street View (GSV) API will be used to collect images from different angles, and a Deep Learning algorithm will be employed to detect the existence of curbs and other accessibility elements in sidewalks. To train this algorithm, a database with street panoramas will be developed and labeled with accessibility attributes. Existing labeled datasets, from other cities, will also be used to train the models and evaluate its generalization. This prototype platform will be tested in downtown Porto to evaluate the accuracy of the model.
Institutions
Main Institutions
- Fundação Ensino e Cultura Fernando Pessoa (FFP)
Other Institutions
- Universidade do Porto Laboratório de Inteligência Artificial e Ciência de Computadores (UP LIACC)
Funding 49.814,20 €
Fundação para a Ciência e a Tecnologia (FCT) - Portugal
49.814,20 €