• Blog
  • /
  • Dive into TECH which helps you with the parking management
on December 30, 2019
  • IoT
  • Machine Learning

An increase in vehicle sales from the past years intensifies a lot of issues due to parking. Finding parking space alone contributes to the burning of barrels of the world’s oil every day. 

Not only in oil but also in congestion of traffic due to parking in road-sides.The number of vehicles (from the data source visualizing) sales shows the scenario in India.

The unmanaged parking facilities make it difficult to accommodate these booming vehicles.

How would this problem be solved?

Now it is necessary to think of a smart parking management system which will be able to

  1. Real-time information on parking space availability.
  2. Booking and canceling of parking space.
  3. Maintaining the vehicle in and out count.
  4. Getting information about unauthorized parking.
  5. Prevent parking on no parking zones.

and a lot more…now insert some tech into this parking management system. 

How can we gather parking space availability data?

We are going to discuss some technologies which could save you time and effectively manage your customers. The solutions data could be extracted using the parking sensor and the IoT-internet of things, or through your surveillance system (CCTV data) by using a combination of machine vision/computer vision and machine learning. Don’t be worried about the jargon words these are not as hard as it sounds. Let’s dive into these two in an abstract:

Machine Vision as Data Source

Computer vision along with the Artificial Intelligent algorithm helps to use the same surveillance system footage installed on the parking premises to identify the unoccupied space. Using these can be any combination of solutions to finding the unoccupied space. Looking into one of the combinations which will be using the object detection library YOLO (create rectangular bounds around the identified objects) or Mask R-CNN (helps to generate bounds and segmentation mask for each instance of an object) and a CNN-Convolutional Neural Network.

Put these into the program, the program catches the picture of a segment of the parking space and, for each parking spot, decides the parking occupancy status utilizing a CNN. The images will be filtered by a mask that will be filtering the unoccupied space. The parking space will be segmented and these will be identified based on the mask and these are classified using CNN which corresponds to the status of being occupied or unoccupied. This occupancy data will be passed to a database..

IoT-Internet of things as Data Source

Using IoT as a base for the data source for parking management requires hardware such as sensors and gateways. Here the sensors identify the occupancy and will be sending this data to a gateway and the gateway routes this data to cloud servers which can be accessed using different integration methods like HTTP, MQTT, etc..

There are several IoT technologies that are being used for this such as LoraWan, SigFox, and many more…

The data from the source which are sent to the cloud can be integrated into any type of application that helps in identifying the parking space, which contributes to more customer satisfaction and less amount of environmental pollution due to the burning oils.

Written By
Yethin Dias


Notify of