Deep Learning-Based Traffic Characterization for Smart Cities (Phase 2)

This project addresses the need to enhance environmental sustainability within the framework of Smart Cities (SC) and Sustainable Cities (CS) by applying innovative technologies, specifically Deep Learning (DL). The core problem involves the efficient and quantitative measurement of urban mobility indicators, such as vehicle counting, identification of parking spaces, pedestrian crossings, and bike lanes. The proposed solution leverages DL for the precise identification, classification, and localization of these elements within images captured around the UPS campus. The methodology encompasses a state-of-the-art study, literature review, image acquisition via drone photogrammetry, map generation using image stitching, exhaustive labeling of transportation elements, and the subsequent training and validation of a DL-based identification model. The expected outcome is the acquisition of reliable quantitative data for measuring transport parameters, thereby contributing to improved urban mobility management.<br/><br/><b>Goal</b>: <br/>The main objective of this project is to apply Deep Learning techniques for the identification and classification of transportation-related elements in the vicinity of the UPS, in order to measure key urban mobility indicators.<br/><br/><b>Research lines</b>: <br/>Technologies applied to natural resources