RRJ4

Reference:
Markoff, J(2010, February 22) Computers Turn Flat Photos Into 3-D Buildings New York Times Retrieved February 24, 2010


Summary:

The article talks about a new system which has been created by the computer science researcher at the University of Washington and Cornell University that could combine the others works and create three-dimensional renderings. This system was named PhotoCity; the creator named Noah Snavely, a computer scientist. He set up algorithms that generated three-dimensional models from two-dimensional photos which were unstructured. The original project came from the Microsoft’s Photosynth service, but the Photosynth was limited to hundreds photos mostly. This technology is able to process many digital photos of one object quickly, and render it in three dimensions. In order to get highly quality, the researcher tries to use the ways like integrating the system to the games, and the game quickly attracted lots of loyal players. The volunteers created a web site to collect photos. Their goal is to create a no boundaries game that can expand to fill the world.

Reaction:

Flash is a kind of famous image processing software. The fundamental idea is that the software collects many pictures which describe the same object in different ways and then makes it look like the object is moving. The PhotoCity has the same theory, but it used the two-dimensional photos to create three-dimensional renderings with some difficult algorithms. But in the three dimensional we need to consider about the different view degrees for the object, and it requires lots of pictures. The author mentioned that it can be collected online, here I have some questions; I think it is too hard to take pictures of one object in a steady space, and even though we can get lots of pictures from the website, it should be only a few photos that focus on the same object.

When we take pictures, we need to consider about a lot of elements around it: light, degree, color, focal length and environment factors like wind and background. If we want to make a three-dimensional rendering for an object, it will be hard to take many pictures unless we have a functional camera. Maybe we need a camera that can take high quality pictures in one second. It will be difficult for the fresh man who just learned how to take pictures. And another problem is lack of the pictures that have similar object and background inside. For example, if we take pictures of a flower, maybe we take it in the garden, or in the park, or the flower has different shape, bigger or smaller, brightness or darkness. So I thought it could not support much help for the three-dimensional creating by downloading the pictures from the website.

In another view, if we really can exploit the system completely, that could be such big technology for human. Like when we want to salvage the sinking boat, we need to consider the landform or the broken degree of the boat. With the help of this system we can easily know the environment of the sinking boat. Actually we can create a three-dimensional view of an object currently, but if we need to consider the environmental elements, we still need to take pictures from the original objects in different degrees. So I thought, no matter what kind of three-dimensional rendering we want to make, a high quality, functional and multipurpose camera is necessary.

When I was little I had never thought about whether we could observe a bird from up and down. Three-dimensional technology can help us to reach the dream. Science is developing rapidly. I hope we can find another way besides the camera to make three dimensionals, even if we don’t take photos directly from the object but still can observe it in different degrees, like the machine in some science fiction movies.