Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
FindMySpace: A Cost Effective Carbon Reduction Solution for Parking
Parking is a major worldwide problem. Emissions generated by drivers searching for parking contribute to the global climate crisis. The goal of my work is to solve the carbon emissions problem created by drivers' search for parking. I have achieved this by building a hybrid deep learning neural network and edge device solution that gives consumers real-time availability of parking and predicts future availability while also being extremely cheap and easy to implement in societies worldwide. I have created a convolutional deep-learning neural network that uses classification to determine whether a single parking space is free or busy. With this tool, I can take a picture of a parking lot and send each space to my model to know how many spaces are free or busy. I also built a smart camera using a raspberry pi and built a prototype app to showcase how the full solution can be used. I have vigorously tested my model with data sets containing individually cropped spaces, videos of parking lots online, and a real parking lot. My solution demonstrated accuracies from 90-97% for all of these, proving that my model can be instrumental in solving the parking crisis worldwide with easy accessibility and low costs.
FindMySpace: A Cost Effective Carbon Reduction Solution for Parking
Parking is a major worldwide problem. Emissions generated by drivers searching for parking contribute to the global climate crisis. The goal of my work is to solve the carbon emissions problem created by drivers' search for parking. I have achieved this by building a hybrid deep learning neural network and edge device solution that gives consumers real-time availability of parking and predicts future availability while also being extremely cheap and easy to implement in societies worldwide. I have created a convolutional deep-learning neural network that uses classification to determine whether a single parking space is free or busy. With this tool, I can take a picture of a parking lot and send each space to my model to know how many spaces are free or busy. I also built a smart camera using a raspberry pi and built a prototype app to showcase how the full solution can be used. I have vigorously tested my model with data sets containing individually cropped spaces, videos of parking lots online, and a real parking lot. My solution demonstrated accuracies from 90-97% for all of these, proving that my model can be instrumental in solving the parking crisis worldwide with easy accessibility and low costs.
FindMySpace: A Cost Effective Carbon Reduction Solution for Parking
Arora, Krish (Autor:in)
24.09.2023
862443 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Cost-effective parking enforcement service
| British Library Online Contents | 1994
British Library Online Contents | 2018
Low-cost concrete parking garage
Engineering Index Backfile | 1955
| TIBKAT | 2005