Enhancement of infrared image using fractal based human visual system

Yu Tian-He,Dai Jing-Min,Kang Wei-Min
DOI: https://doi.org/10.1117/12.821232
2009-01-01
Abstract:This paper presents a fractal based human visual system(HVS) proposed for elimination of blurred edge and enhancement of infrared images. Human visual system is used in the proposed approach and fractal dimension of an infrared image is first measured using a fraction Brownian motion model as an accurate indication of the coarseness of infrared image. The fractal dimension of each pixel was calculated using a fractional Brownian random model and the self-similitude and surface coarseness of the gray level of each pixel are described using fractal value of each pixel. HVS is sensitive to the edge area of an image, and the pixels are then classified into edge pixels, smooth pixels and texture pixels according to their fractal dimensions. The gray level of each edge pixel is weighted and enhanced using a fractal based HVS. Experimental results indicate that fractional Brownian motion model is effective in detecting the edge detail of an infrared image. This approach can be used to preserve edge details and enhance the contrast of an image.
What problem does this paper attempt to address?