What Is Computer Vision?

Daniel Liden

Computer vision is a field of study and research in computer science and engineering that focuses on computers and machines that can receive and interpret visual data. The concerns of this field can be as simple as devising and integrating cameras that work well with computers or as complex as developing visual systems that enable computer technologies to interact with users. While there are many different potential applications for computer vision, medical technology has become one of the most practical and accessible fields for the implementation of such visual technology. Highly detailed images of patients can provide a great deal of valuable diagnostic data that leads to highly personalized and useful medical data.

Engineers use computer vision to design automated quality control mechanisms in manufacturing.
Engineers use computer vision to design automated quality control mechanisms in manufacturing.

There are many different elements of computer vision that often must be combined to make a cohesive and useful vision system. In cases that require anything more complicated than reproducing an image, for instance, some level of image recognition or detection is usually required. Computer vision technology is designed to recognize specific visual cues, such as those on human faces, in order to focus on or track a given object. Some technologies are designed to recognize text, often with the purpose of "translating" the text from an image file to a text file that can be edited and manipulated.

Computer vision technology may involve nothing more than a video camera connected to a computer containing interpretive software.
Computer vision technology may involve nothing more than a video camera connected to a computer containing interpretive software.

Computer vision is commonly studied in conjunction with biological vision, the process by which organisms such as humans receive and interpret visual data. The two fields of study contribute significantly to each other. Advances and developments in computer vision can suggest possible mechanisms by which biological vision occurs. Discoveries in biological vision, on the other hand, can provide ideas for new ways for computer technology to handle external visual data. It is not uncommon for biologists, computer scientists, and engineers to work together on projects regarding computer or biological vision.

There are many different fields, most within the sciences, that make regular use of computer vision technology, usually for research purposes. Artificial intelligence, a common area of study in computer science and engineering, uses visual technology to devise navigation or recognition systems for robotics. Computer vision technology is sometimes used in optics because artificial visual systems can be made to "see" and record a wider range of visual data than organic visual systems can. Additionally, many different fields contribute to the development and implementation of visual technology in computerized systems. Mathematics, for instance, is an essential element of the programming that goes into the interpretation of visual data by computers.

You might also Like

Discussion Comments


@miriam98 - I agree that the human eye is still more advanced than any comparable computer technology. However, robots can do something that human eyes cannot – they can discern the infrared spectrum.

Actually, I’ve heard that you don’t even need a robot’s “eye” to do this; you can retrofit some of the cheaper digital camcorders so that they can deliver true night vision, or a buy a camera with that technology built in.

Once robots are unleashed upon society, they will have powers of perception that surpass anything that the human eye possesses.


@hamje32 - The marriage of computer vision and biological vision is brilliant. There are actually devices on the market that enable people who are visually impaired to achieve near perfect vision.

Perhaps someday it will be possible to use computer implant technology to cure blindness. It makes you gain a newfound appreciation for how powerful and how wonderful the human eye is; its ability to survey the landscape around us and discern a multitude of colors with a breadth of detail.

Computer vision is not quite there yet, but I believe that it’s catching up.


@Charred - I am fascinated by the prospects for computer vision systems in medical imaging and analysis. I think here pattern recognition might be useful.

Medical images contain huge amounts of information, and I think it would be easy for a mere human to overlook something that, however slight, was significant.

Computers are good at number crunching large quantities of data. I’m sure they would be able to do the same with medical imaging; except of course, it would be spatial data. I don’t know how that would work, but evidently it does.


We dealt with image analysis algorithms when I took my computer programming classes in college. One of the most popular of these is video motion detection, which would be used in tracking computer vision applications.

The premise is simple: how do you detect motion in a video image? The answer, in one form or another, is to compare frames of images with each other. You might take each frame of an image and compare it with the prior frame.

If there is substantial difference (defined as a certain threshold) then the computer figures out that motion has taken place. These kinds of things are sometimes incorporated into security cameras and stuff like that, but it’s not hard to build a simple motion tracking application. You can find lots of code and sample applications on the Internet.

Post your comments
Forgot password?