ICIP 2006, Atlanta, GA
 

Slide Show

Atlanta Conv. & Vis. Bureau

 

Technical Program

Paper Detail

Paper:WP-P1.10
Session:Visual Object/Event Detection, Segmentation, and Classification
Time:Wednesday, October 11, 14:20 - 17:00
Presentation: Poster
Title: AN NEW BOX-COUNTING METHOD FOR ESTIMATION OF IMAGE FRACTAL DIMENSION
Authors: Jian Li; Chongqing University 
 Caixin Sun; Chongqing University 
 Qian Du; Mississippi State University 
Abstract: Fractal dimension (FD) is an effective measure for complex objects. It is widely applied in the fields of image segmentation and shape recognition. This paper presents a new box-counting method for estimation of fractal dimension of images. Original ideas of the method came from the principles of Differential box-counting (DBC) method. Improvements appear in the approach on partition of image plane and computation of the number of boxes covering the image. The computation results of fractal dimension of simulated images and Brodatz texture images demonstrate the new method is effective and produces a wider range in fractal dimension than the original DBC method.