ParkSight

by Faiyad Ahmed Masnoon

Github Logic Website

Hover over each card to see more details.

Please switch to a computer for a better viewing experience.
This site is not optimized for smaller screens.

Best Spot Algorithm Please wait for file to load. If you are on mobile, please switch to a computer.

Best Spot Algorithm

Normal behaviour of the algorithm

Finds the most optimal free parking spot, talking into account the distance from the user and the adjacent parking spots' status.

Real-time Switching Please wait for file to load. If you are on mobile, please switch to a computer.

Real-time Switching

When a parking spot frees up, the algorithm automatically recalculates the scores for the free spot, and if it is a better choice, directs the user to the new spot.

Prioritizing Space Please wait for file to load. If you are on mobile, please switch to a computer.

Prioritizing Space

To ensure a smoother parking experience

If there is a parking spot with 2 adjacent spots free, it will be preferred over a parking spot with 1 adjacent spot free, provided that it is not too far away. Simillarly, a spot with 1 free adjacent space will be preferred over a spot with none.

Precision

The algorithm is precise, as seen here: first it choses the edge spot, even though it is further away since it is more optimal. Then it switches to a spot with a free adjacent spot.

Logic

This program uses a Simple Vector Classifier to identify empty parking spots. It then applies a score-based algorithm that considers the position of each parking spot, its distance from the user, and the status of the spot itself as well as its adjacent spots to determine the most optimal parking location. Once the best spot is identified, the system directs the user to it, dynamically updating its recommendation in response to real-time video footage.