The iphones built-in camera is unable to compete with dedicated cameras for image quality. However, you have it with you everywhere you go, and chances are good that you dont carry your camera except to special events. Therefore, there are many times when you want to milk the last bit of image quality out of your camera.
Two areas where you may want to enhance your iphone camera are:
1. Resolution and field of view
2. Low light conditions (night shots)
Resolution and Field of View enhancement
A 2-megapixel resolution isnt too shabby for a tiny camera hiding in the corner of your iphone. if you where you want a much bigger image, you may be missing your 8-megapixel dedicated camera.
The solution is to stitch your photos together to create a larger frame. one common application for this is when you want to photograph a large poster at a conference, so you can read the text later, but the fonts are too small to see in one frame.
If you have a Mac, you can use software called calico panorama, manufactured by kekus Digital (www.kekus.com). it is relatively cheap: $39 as of this printing. on a pc, you can use freeware software called autostitch that uses a prior version of the same code engine (www.cs.ubc.ca/~mbrown/autostitch/autostitch.html).
First, take a panorama set of photos, that is, a series of photos, making sure to overlap them generously. When going from side to side, i ind it useful to move the camera just enough to put the right half of the screen onto the left half. Therefore, for a total image width of three ields of view, you will take ive side-by-side photos. if you wish, you can do multiple rows as well.
The result is the maximum ield of view that the software can produce from the images you gave it. It will automatically align them and adjust contrast and color settings to make the result seamless.
Low Light conditions
The iphone camera is pretty smart about auto-adjusting contrast when bright and dim objects are both in the ield of view. nevertheless, it requires a minimum light intensity in order to form an image. at night, for example, there often is insuficient light to create a nice photo. however, if you can distinguish any shapes in your image, then you can merge dozens of photos in order to produce a single higher-contrast image. You need to have one well-deined visible object in the image.
This can be a penlight, street lamp, or whatever. This provides a single object that is visible in each image and allows you to align the images. next, you will need an open source software package available (for Macs only!) called Lynkeos.
Generally, it is used by astronomers to improve the quality of images taken through telescopesbut theres no reason that you cant use it to create images of the beautiful city lights taken from your drive home after skiing (when you had only your iphone).
First, you need to obtain images. set your camera against an object so that it is very still. even a slight shake of your hand will introduce blurring in the inal image. This means you can set your iphone on its side on a rock, table, against a tree, or something like that. You can even use a shoe as a tripod. shoes and socks are handy because you (hopefully) always have them with you and they wont scratch your iphone.
Adjust your iphone position so that you can both reach the snapshot button and so that your frame contains at least one sharp, bright object. You should see it well in your camera roll on every photo.
This is critical or Lynkeos will not be able to squarely align the images.
Take as many images as you can standyou will probably need several dozen, and you wont know if you got enough until you go to process them. The more images you take, the better your inal image becomesassuming you are photographing a still environment. Unfortunately, this technique wont work for catching a moving object! (no photographing nocturnal panthers using
your iphone!)
When you get back home, download the images onto your computer, and use Lynkeos (http://lynkeos.sourceforge.net) to stack the images and produce one final high-quality frame.
<>
January 14th, 2010
fix
Posted in
Tags: 
