Adaptive linear pixel-based fusion for contrast enhancement
Cheng‐Hsiung Hsieh, Bo-Chang Chen, Qiangfu Zhao · 2012
This paper presents an adaptive linear pixel-based fusion for contrast enhancement which is abbreviated as ALPS. The proposed ALPS approach combines the histogram equalization (HE) and image fusion schemes into one. A well-known problem in the HE is over-enhancement. It is also noted that a pixel-based fusion may suffer the problem of anomaly pixels. To avoid the problem of over-enhancement in the HE, an adaptive pixel-based fusion scheme is applied to the original image and its equalized image by the HE. Moreover, to deal with the problem of anomaly pixels, two restrictions are posed and a low-pass filter is used in the proposed ALPS approach. The simulation results indicate that the over-enhancement is relieved and that better enhanced images are for the proposed ALPS approach when compared with a HE-based contrast enhancement scheme. Also, the problem of anomaly pixels can be avoided by the proposed ALPS approach which is verified by the given examples, while the compared pixel-based fusion approach suffers it.