Machine perception is the capability that allows the computer system to perceive the environment in a similar way to humans. Biological system uses sight, touch, smell, sound and taste to process data coming from the environment. The field that deals with machine’s senses is a part of Artificial Intelligence, or AI, which has a goal to produce computer software with intelligent behaviour.
Computer vision includes methods for collecting, analysing and understanding images and translating them into numerical information. The device with an ability to perform tasks based on data received from the environment can be used for a variety of purposes, ranging from face recognition to space exploration and military.
The everyday use of sight made humans forget the complexity which lies behind it. People will easily recognize some pattern in the image, but the problem occurs when it needs to be explained to the computer.
There are two main approaches to pattern perception: mathematical pattern recognition and descriptive techniques. The tasks that machines perform usually include object recognition, motion analysis, scene recognition and image restoration.
This research area uses findings from different branches of science: solid-state physics which explains optics, the core part of the vision system, or neurobiology, used to develop reactions similar to the conscious being. It also uses mathematics and geometry to process and analyse images.
Psychology has been a special area of interest too. Numerous psychological results and insights helped to understand perception and work on developing one for the machine. However, psychology didn’t provide many models that can be directly used in machine programming. For this field, more useful was knowledge about the ways eyes receive signals rather than how the brain understands them.
Nevertheless, the progress in machine vision has significantly improved in recent years and AI’s benefit to humans is being increasingly recognized.
“Never send a human to do a machine’s job.” – Agent Smith, The Matrix (1999)
Machines can replace humans in performing simple tasks such as mounting in factories, but they could also make, and already are making, a huge contribution to humans’ understanding of space and sea as well improve different aspects of human life.