Computer Vision


We have reached the stage where the information collected is examined in detail across multiple dimensions. Now, we can distinguish the parts that make up the whole, see if there are repetitive patterns, decide what the general theme is, and identify errors that should not even exist. Examples of applications based on analysis capabilities of computer vision are security management 1, license plate recognition 2, workplace security 3, line cameras in sports 4 and detection of fake products 5.

1 Public Safety

For safety reasons, there are plenty of cameras in subways. In multi-story and multi-line stations, incoming images are continuously analyzed by the computer to ensure accurate and fast intervention. Often the aim is to detect people who are fighting or pushing each other rather than those who are and alert security officials’ attention on relevant cameras.


2 Vehicle Detection

When it first came out, vehicle detection was difficult, expensive and inaccessible. Today, with the rapid development of smart cameras with internet, they escalate efficiency in various areas such as security or road toll. The system works with the camera and the software within the camera or in the main computer. Instant imaging of the license plate is matched with the data in the mainframe. Without issuing any ticket, it is now possible to register cars within parking lots by matching the entry hour and the license plate. It has long been a part of our lives to track traffic violations on highway via cameras. Likewise, searching for stolen or surveilled cars not only by license plate but also their brand and model information has become one of the factors that increase public safety.


3 Workplace Security

Occupational accidents are one of the undesirable consequences that may happen not only in heavy work environments but also in hospitals, warehouses or ordinary offices. With an application combined with computer vision and artificial intelligence, Microsoft presented a good example of how technology can improve workplace safety. Ensuring that employees are wearing their helmets, an alert being sent to the surroundings after a hazardous barrel being hit by a forklift or warning a nurse about a patient who is walking more than necessary after a heavy operation can be possible by computer vision and simple cameras.


4 Video Assisted Referee (VAR)

After tennis and volleyball, cameras on the bylines and at the back of the goal have started to be used actively in soccer as well. Video Assisted Referee (VAR) records controversial positions whether the ball has crossed the line or players are offside. The interpretation is currently done by humans but it is possible to predict that this task will be performed entirely by computer vision in the near future.


5 Detecting Fake Products

All over the world, it is very common to make fakes of expensive and popular products. It is often a serious problem to identify fakes which are easily sold at an affordable price on online shopping sites. Products supported by computer vision to identify fakes have gradually begun to emerge. Thanks to data processing methods to examine fine details that the naked eye cannot notice such as texture, stitching, folding, we will be able to understand with 99% accuracy which product is real or fake.