Mercurial > cortex

     1 * Object Recognition from Local Scale-Invariant Features, David G. Lowe

     2   

     3   This is the famous SIFT paper that is mentioned everywhere.

     4 

     5   This is a way to find objects in images given an image of that

     6   object. It is moderately risistant to variations in the sample image

     7   and the target image. Basically, this is a fancy way of picking out

     8   a test pattern embedded in a larger pattern. It would fail to learn

     9   anything resembling object categories, for instance. Usefull concept

    10   is the idea of storing the local scale and rotation of each feature

    11   as it is extracted from the image, then checking to make sure that

    12   proposed matches all more-or-less agree on shift, rotation, scale,

    13   etc.  Another good idea is to use points instead of edges, since

    14   they seem more robust.

    15 

    16 ** References:

    17  - Basri, Ronen, and David. W. Jacobs, “Recognition using region

    18   correspondences,” International Journal of Computer Vision, 25, 2

    19   (1996), pp. 141–162.

    20   

    21  - Edelman, Shimon, Nathan Intrator, and Tomaso Poggio, “Complex

    22   cells and object recognition,” Unpublished Manuscript, preprint at

    23   http://www.ai.mit.edu/edelman/mirror/nips97.ps.Z

    24   

    25  - Lindeberg, Tony, “Detecting salient blob-like image structures

    26   and their scales with a scale-space primal sketch: a method for

    27   focus-of-attention,” International Journal of Computer Vision, 11, 3

    28   (1993), pp. 283–318.

    29   

    30  - Murase, Hiroshi, and Shree K. Nayar, “Visual learning and

    31   recognition of 3-D objects from appearance,” International Journal

    32   of Computer Vision, 14, 1 (1995), pp. 5–24.

    33 

    34  - Ohba, Kohtaro, and Katsushi Ikeuchi, “Detectability, uniqueness,

    35   and reliability of eigen windows for stable verification of

    36   partially occluded objects,” IEEE Trans. on Pattern Analysis and

    37   Machine Intelligence, 19, 9 (1997), pp. 1043–48.

    38 

    39  - Zhang, Z., R. Deriche, O. Faugeras, Q.T. Luong, “A robust

    40   technique for matching two uncalibrated images through the recovery

    41   of the unknown epipolar geometry,” Artificial In- telligence, 78,

    42   (1995), pp. 87-119.

    43 

    44     

    45      

    46 

    47 

    48     

    49 

    50
author	Robert McIntyre <rlm@mit.edu>
date	Sun, 10 Mar 2013 18:17:53 +0000
parents
children	9c37a55e1cd2