Higuchi Fractal Dimension with a multidimensional approach for color images

Author(s)

Tenguam, Jaqueline Junko ; Longo, Leonardo H. da Costa ; Roberto, Guilherme Freire ; Tosta, Thaina A. A. ; Silva, Adriano B. ; do Nascimento, Marcelo Zanchetta ; Neves, Leandro Alves

• Document type: Journal article
• Source: SOFTWARE IMPACTS; v. 21, p. 6-pg., 2024-08-06.

Abstract

Among the methods for estimating fractal dimension, the Higuchi approach is limited to processing grayscale images. To improve this approach and estimate Higuchi fractal dimension values from color images, multidimensional and multiscale Higuchi software was developed for the pattern analysis present in images considering the RGB color model. The software can provide local/global fractal dimensions. Moreover, it was tested to quantify images from important datasets commonly explored to evaluate computer-aided diagnosis schemes (histological images of colorectal cancer, liver tissue, and oral epithelial dysplasia). The results were relevant and useful for the process of pattern recognition considering the investigated groups. (AU)

• FAPESP's process: 22/03020-1 - Normalization of H&E stain by autoencoders with analysis of ensemble learning for histological images
• Grantee: Thaína Aparecida Azevedo Tosta
• Support Opportunities: Research Grants - Initial Project