Introduction: DIY Camera Array 1: Computational Photography Primer.
DIY Camera Array 1: Computational Photography Primer.
DIY Camera Array 2: Computational Refocusing With Just One Camera
DIY Camera Array 3: ... TBA
This first Instructable is a primer on the field of Computational Photography, which is a new field of research that is developing extremely powerful cameras. These cameras allow the Depth of Field, the object in focus, and the position of the camera to be modified after the picture is taken. None of those things are possible with a traditional camera.
We show how to build one kind of computational camera -- a light field array. There are many other designs out there.
Imagine focusing a picture after you take it.
Imagine focusing right through fences, trees, and people.
Step 1: What Is It All About?
Over the last half decade, a new field of research has developed, called Computational Photography. Computational Photography is still in the process of defining itself. As a young field, it incorporates insights from many other fields: mathematics, image processing, optics, art, visual neuroscience, and more. Central to the field is extending the capabilities of the camera. You may be surprised to learn that cameras have hardly deviated from the camera obscura (box-with-a-hole-in-it) design over the last hundred fifty years. Computational photography is, to date, the most radical re-imagining of what a camera is, and what a camera can do. Frédo Durand recently said "Computation is the new Optics" and this is probably the best field-defining phrase we've heard yet. But there is no computation without hardware.
Researchers have created many new camera designs, extending and enhancing almost every functional element of the camera. Many of these new cameras attempt to capture a richer representation of light, which is called the "light field". What is the light field? Surrounding you, now and always, is a reverberating volume of light. Rays of light bounce around you from every possible angle. Some new computational cameras attempt to capture this light field, which has the potential to give the photographer unprecedented creative control... after taking the picture.
Step 2: Wrapping Up the Primer... Be Sure to Read Part 2.
This camera array, with its twelve cameras, captures only a very coarse representation of the light field. It is far from the state of the art. However, many of the most sophisticated light field cameras are confined to the laboratory, or cost from $5-10,000. This camera array (from here on, called the Large Light Field Camera Array or LLFC) can be built for a few hundred dollars with old used cameras, and is happy to go outside and take pictures.
The refocusing technique demonstrated here is called Synthetic Aperture Refocusing. Ted Adelson at MIT wrote an early influential paper on it, and it was also investigated in depth by many people at many labs, including MIT (Frédo Durand, Ramesh Raskar, Ren Ng) and Stanford (Mark Levoy). Todor Georgiev at Adobe has developed and documented some very sophisticated camera systems and written up lots of tutorial material. Synthetic aperture refocusing is now pretty well understood. Our camera array may be the first to exploit cheap, high resolution cameras that don't need a computer connected, and our refocusing software LFtextures may be the first open-source, cross-platform application available. Our goal is to make these technologies accessible and understandable, and to encourage experimentation by people of all stripes.
This tutorial is just the beginning. We (Daniel Reetz and Matti Kariluoma) will present a series of tutorials on how to build an array, how to operate it, and how to make synthetically refocused images from the output of the array. These updates will come about 1 per week for the next six or seven weeks. While this deviates a bit from the normal Instructables format, we felt it would be a better approach than a hundred-step Instructable, as I did with my previous "DIY Book Scanner" project. PS. If you're interested in book scanning, drop by the DIYBookScanner forums.