Mobile Edge Computing for Rapid deployment Object Detection System

In the past, the process of transferring data collected by IoT devices or mobile devices has data leakage, personal privacy, and information security issues since object detection systems have mostly sent data to the cloud for processing and storage. To solve this problem, we implement object recognition and tracking function in JavaScript on the front end, and set up the CRUD operation API in Python on the back end, so that users can directly perform object recognition on the computer or mobile browser, and manage the database on the web through the API. It enables the process from deployment to data transfer in a very short time.

I. Introduction

The traditional parking lot is often stationed in the booth to control the vehicle in and out, this way may have the problem of increased personnel costs. Image recognition can be applied to parking lot management, saving school personnel costs, making management more convenient, and making it easier for students to query parking spaces, achieving the effect of a smart campus. Traffic flow recognition will make the cost of parking management greatly reduced.