The adaptive contrast enhancement (ACE) algorithm is a widely used image enhancement method, which needs a contrast gain to adjust high frequency components of an image. In the literature, the gain is usually inversely proportional to the local standard deviation (LSD) or is a constant. But these cause two problems in practical applications, i.e., noise overenhancement and ringing artifact. In this paper a new gain is developed based on Hunt's Gaussian image model to prevent the two defects. The new gain is a nonlinear function of LSD and has the desired characteristic emphasizing the LSD regions in which details are concentrated. We have applied the new ACE algorithm to chest x-ray images and the simulations show the effectiveness of the proposed algorithm.
|Number of pages||5|
|State||Published - 2 Nov 1996|
|Event||Proceedings of the 1996 IEEE Nuclear Science Symposium. Part 1 (of 3) - Anaheim, CA, USA|
Duration: 2 Nov 1996 → 9 Nov 1996
|Conference||Proceedings of the 1996 IEEE Nuclear Science Symposium. Part 1 (of 3)|
|City||Anaheim, CA, USA|
|Period||2/11/96 → 9/11/96|