Learn something new every day
More Info... by email
Computer vision, also known as "machine vision," is a technology that uses cameras and computers to interpret images. There are many different uses for this technology. Some of the most common computer vision applications are in the medical, industrial, and security fields. Additionally, machine vision is prominent in robotics.
Each application of machine vision has the goal of gathering useful information based on visual clues. The data used in computer vision applications may be from a static source, such as a photograph. This technology can also be used to interpret moving images, including live or pre-recorded action captured through a video camera.
Medical computer vision applications are typically used to process static images. Microscope results, x-ray pictures, and ultrasound images can all be interpreted by this technology. Vision software can be programmed to detect abnormalities in a medical photograph. Computer analysis, for instance, can be used to locate tumors on an x-ray result. Computers are sometimes able to scan medical images and identify potential problems at a faster rate than human technicians.
Industrial applications can also make use of machine vision. Factories often use computer vision to inspect merchandise for defects, or to sort objects based on attributes such as size and color. Some factories use high-resolution cameras to capture extremely detailed images of products. Vision software is then used to automatically locate small fractures or imperfections in the material. This technology is able to view details that are imperceptible to the naked eye.
There are several computer vision applications within the field of security. Computers are able to analyze live video feeds in order to track important patterns. Security checkpoints at airports, for example, sometimes use machine vision to recognize the faces of previously identified, wanted criminals. Vision software is also able to track individuals in a crowd and identify suspicious activity, such as abandoned baggage or loitering.
Robotic systems frequently employ computer vision. Autonomous vehicles, including unmanned aerial vehicles (UAVs) and lunar rovers, often use cameras and computers to analyze the nearby landscape. Prominent terrain features such as mountains can be compared to an electronic map. This allows robotic vehicles to determine their location based on external reference points.
Computer vision is an emerging technology that has not yet reached its full potential. Many scientists believe that in the future, machine vision will lead to advanced technological breakthroughs. Potential applications may be applied to automated cars, unmanned airliners or other technological advances.
@nony - I think that level of image analysis would be taxing for a computer, I agree.
I think the most practical uses of computer vision is in anything that needs to detect body heat. I believe that some of the advanced motion detection sensors use this.
The cameras look at the image using infrared or something like that; when humans enter the image frame, the heat becomes visible as red in the frame and that trips the sensor.
@MrMoody - I heard someone talk about a unique application for vision software once – website filters.
Basically the software would filter for those images that, frankly, look like nothing but human flesh; adult sites in other words. If it detected that what it was looking at was all flesh, the filter would block the site.
As a matter of fact I seem to recall Google or one of those other companies discussing the possibility of creating such a sophisticated filter.
The problem, they explained, is that human flesh comes in wide varieties of tones and colors. So those individual numbers that you described can run the gamut from very low values to very high values, with potentially millions of possible combinations.
The raw processing power needed to conduct such an analysis would make it impractical however to integrate into a web browser, at least for now anyway.
I am software developer by trade and have some basic familiarity with the idea of image analysis. Basically a computer can scan an image for certain patterns to determine if there are abnormalities or other things that stick out.
It does this by giving a numerical value to each pixel in the image and then treating the whole image as a statistical set. You can have parts of the image that seem to have more of one kind of value than another.
If you have a region like this, and then all of the sudden a value that doesn’t “fit in” appears in the image, the program will zero in on that part of the image for further analysis. That’s an oversimplification but that’s the basic idea.
All images are just rows and columns of numbers and can be evaluated statistically like any other set of numbers.
One of our editors will review your suggestion and make changes if warranted. Note that depending on the number of suggestions we receive, this can take anywhere from a few hours to a few days. Thank you for helping to improve wiseGEEK!