iPhone Captures 4MP Images

As we all know, iPhone camera is a 2 megapixel camera. How could it capture 4 megapixel images? A recent iPhone application, ClearCam, can help you. ClearCam provides iPhone the enhanced capturing mode, by taking 6 photos sequentially in a very short time (2.5 seconds). It then creates a sharper, 4 megapixel image with less noise by combining these 6 photos. 

The technique used here is called multi-frame image super-resolution in computer vision. The idea is to increase the resolution of a captured image by fusing information from multiple images or a short video.   
 

Technology

The key challenge for multi-frame image super-resolution is image alignment or image registration, i.e., all the images should be aligned with a reference image. The shift between the reference image and any other image should have sub-pixel accuracy, such that the super-resolved image reconstruction is possible to contain more information. The reference image could be the first frame, or the best frame in the pool. It often assumes an affine transformation model, which involves translations, rotations, and scales. 

This is a cool iPhone application using computer vision technology. The following figure illustrates one of the impressive results using ClearCam. 

What is the processing speed? According to iSmashPhone.com’s interview with ClearCam, it typically takes 30-40 seconds for an enhancement.

If you are interested in learning more about this technology, I would recommend reading the following references by Dr. Shmuel Peleg:

  1. A. Rav-Acha, A. Zomet, S. Peleg, Robust Super Resolution, CVPR’01, Hawaii, Dec. 2001, Vol. I, pp. 645-650.
  2. A. Zomet and S. Peleg, Super-Resolution from Multiple Images having Arbitrary Mutual Motion, in S. Chaudhuri (Editor), Super-Resolution Imaging, Kluwer Academic, Sept.2001, pp. 195-209.

 

No Comments

Be the first to comment!

RSS feed for comments on this post. TrackBack URI

Leave a comment