Smart Parking Inspection System

Using AI to Revolutionize Parking Management

Developed by Nyabiosi Sydiney Nyabiosi - 2024 Capstone Project

Efficient parking space management is crucial for urban traffic flow. However, traditional methods often rely on manual inspections, resulting in inaccuracies and increased workload. To address this challenge, this project focuses on developing a Smart Parking Inspection System.

The system's core functionality revolves around parking permit verification, utilizing Optical Character Recognition (OCR) technology to read and validate vehicle registration plates in real-time. Additionally, it incorporates an alert mechanism for prompt notification of permit violations and features a database system to store all parking and vehicle information.

Nyabiosi Sydiney Nyabiosi

Train Model

Using a dataset comprising around 150-200 images of cars with clearly visible number plates, we fine-tune a pre-trained Mask R-CNN model. This enables accurate detection and segmentation of number plates.

View Training Process

Image Analysis

During inference, the trained model processes input images to detect number plates and read their values. It outputs the coordinates of bounding boxes around detected cars and their license plate information.

Try Image Analysis

Video Analysis

By applying the detection and recognition pipeline to each frame of a video stream (live feed or saved video), we can continuously monitor and analyze parking lots in real-time, enabling efficient parking management.

Explore Video Analysis