Present

Computer Vision is such a broad category that people are bound to continue working on projects that center around the topic.  The Penn State Computer Vision Lab, for instance, wanted to supplement speech recognition with additional ways to recognize communication signals.  They used computer vision to attempt to add the ability to recognize gestures to the ability to recognize language.

In addition to this project, Penn State also attempted to clarify unclear text.  They had a project that would recognize and clarify text in images using computer vision so that the computer could read it and use it.  An example of this is below, an image of a stock ticker in New York.  The text is not totally clear, and this project that Penn State did could attempt to fix it.

The University of California-Berkeley is also working on projects having to do with computer vision.  The first project has to do with attempting to remove adult content from websites kids potentially would visit.  Computer vision could detect pictures in which there are human nudes, which would then be off limits to underage people.  A picture having to do with this project is unnecessary.  Berkeley also wants to be able to analyze the content inside images, such as the tiger in grass below.  The computer can see the outline of the tiger, and filled the inside of the outline with orange.  The shape of the grass is imprinted, with the lake analyzed to be blue, and the grass found to be green.

While projects are being done to improve computers, computer vision is already being used by computers today.  The main use of computer vision is optical character recognition (or OCR). OCR attempts to read words using different methods.  It could use computer vision to recognize a word based on the characters in the word, segments of the word, or the shape of the word.  However, as one would expect, OCR is not perfect.  Underlines in the subject text totally confuses computer vision as to what character is being described.  It also gets confused when trying to analyze letters such as 'I', 'i', and 'e'.  The main reason why the 'e' is confused is because when it is distorted on an image, it could look like other characters, such as the letter 'c'.  Yet, OCR is not a total failure, which is why it is being used by software today.  Calera Wordscan Plus 1.0, Caere Omnipage Professional 3.0, and Xerox Imaging Systems AccuText 3.0 already use OCR software.

Of course, it is no surprise that there have already been success stories in the field of computer vision.  However, the success stories that affect average people such as us the most are machine vision.  Machine vision specifically has to do with robots and machines.  It has already been successfully used to detect poor quality in certain manufactured products and can somewhat strengthen security.  In addition, machine vision has been used to track actual movements.  This improves the quality of video games, making the user feel as if the person were actually moving like a real person.

Computational vision has brought about countless success stories.  Unfortunately, these don't benefit society as much as the success stories in the field of machine vision do.  One of the many examples of successes in computer vision is Marr's Theory.  This theory has to do with transforming certain shapes into other kinds of shapes using the information computer vision detects from the image.  An example is seen below.  Given an image, the computer can generalize the image into the shape of a cylinder. The computer can then separate the cylinder into a human-like form with a head, two arms, two legs, and a body.  The arms can then separate themselves into forearms, which, in turn, separates into fingers, creating a hand.  Marr's Theory is not nearly the only success story in the field of computer vision, of course.  There are too many to list in the first place.