Project Background

Markme is an automated attendance system using image/video processing. At the entry gates, high-resolution cameras require installation. Once an employee enters the organization his/her attendance gets marked automatically with a greeting message.


  • 1. Manage employee data.

  • 2. Keep multiple images of employees in the system.

  • 3. Match photo/video of the employee while entering the organization to identify the person and mark the attendance.

  • 4. Keep all the records of multiple entries and exits in a day but, use the first and last impression as the main entry and exit time of employees for attendance.

  • 5. Make it seamless, an employee doesn't need to stop and wait for the system to scan their face.

  • 6. Use multiple cameras feeds to receive data.

  • 7. Announce the name of the person at the first entry, only when he is entering alone.

Technical Stack

  • Python
  • Angular-5
  • MySQL
  • Raspberry pie

Project Screens


  • 1. We installed a camera at every gate and connected it to a single controller like raspberry pie.

  • 2. Once a motion is detected, the cameras become active and starts sending streams to the server. Now, from the server, it gets identified with the help of image processing and face recognition system which confirms that a particular person has entered.

  • 3. All the records of incoming and outgoing appearances are recorded but only the first and last occurrence of the shift is used as signing in and out time.


  • 1. Multiple people enter at the same time was challenging to track and record.

  • 2. Face recognition has to be done utilizing multiple images of a person. An Employer can use multiple images to register the employee. Once an employee is registered, it’s a self-learning model which learns every day and becomes more intelligent.


We have successfully launched the beta version of the attendance system to mark the attendance of 35+ employees.

Get In Touch

Let’s meet and discuss your idea over a cup of filter coffee and we’ll help you make it the next big thing of the decade