Curbside parking for downtown San Jose is becoming increasingly difficult. Even for residents with parking permits, finding spots promptly is challenging. Currently, curbside parking is extremely inconsistent and difficult to predict. And unlike metered parking, curbside spots are not clearly divisible, targetable, and static. By incorporating computer vision and cloud computing, our system, SpartanPark, helps people find elusive parking spots around downtown San Jose, while effectively tackling curbside parking congestion.
SpartanPark consists of a mobile application, image processing servers, and a network of low-energy solar powered cameras with networking capabilities. In the mobile application, on the frontend, users can choose a location on a map to then display an overlay of available spots, within a radius from the chosen location. On the backend, the server will process street images, using curb visibility and colors as indicators for parking availability. For hardware, low power and solar panel equipped cameras will be mounted on street lights and intersections and will potentially be connected to the city’s electrical system through the street light. These cameras will take photos at timed intervals determined by statistical data and will send them to servers for processing. This implementation will be constructionally flexible, easy, and non-obtrusive.
This project combines engineering, computer science, and design disciplines to solve the pressing parking issues plaguing San Jose locals. When applied to its fullest extent, downtown San Jose will experience significantly less traffic congestion, less public lateness, and a more streamlined parking procedure. Further development can allow for customized space detection.