INVESTIGATORS

Dr. Horace H S Ip, Dr. Dinggang Shen, Mr. K T Cheung

 

BRIEF DESCRIPTION

In this project, we have provided a generalized image normalization technique which basically solved all problems in image normalization, including orientation and fold number detection for all types of images. The technique is robust against certain degree of noise. It is shown that any image can be represented by a unique orientation defined by at most three non-zero generalized complex (GC) moments. The correctness of the techniques is demonstrated theoretically as well as in practice by applying them to a number of "degenerate" images which have failed other previously reported techniques for image normalization. This technique is potentially useful for invariant pattern matching in shape recognition and retrieval. We are currently investigating whether the approach can be extended to content-based image indexing and retrieval.

 

FUNDING AGENCY

N/A

 

PUBLICATIONS

1. "Optimal Axes for defining the orientation of shapes", IEE Electronics Letters, 32(20), 1873-1874, Sept. 1996. [Dinggang Shen and Horace H S Ip]
2. "Generalized affine invariant image normalization", IEEE Transactions on PAMI, 19(5):431-440, May 1997. [Dinggang Shen and Horace H S Ip]
3. "Normalization method suitable for every kind of image", ICARCV'96, Singapore, December, 1996. [Dinggang Shen and Horace H S Ip]

FIGURES OR IMAGES