ACE-HF Framework: Enhancing Implicit Information in Hyperspectral Lacquer Coffin Paintings through Automatic Color Enhancement and Homomorphic Filtering

Hyperspectral lacquer coffin painting is an important cultural heritage in China. However, distinguishing polluted areas in visible light is challenging in the visible light due to the influence of pollutants. Hyperspectral imaging, obtained by imaging in tens to hundreds of continuous and finely divided spectral bands, can extract image information beyond the visible light range more easily. Therefore, using hyperspectral data, this paper proposes a method for enhancing implicit information in hyperspectral lacquer Coffin painting based on automatic color enhancement and homomorphic filtering (ACE-HF) framework, which combines histogram matching, automatic color enhancement (ACE), and homomorphic filtering. Firstly, ACEHF framework uses minimum noise fraction (MNF) denoising to obtain the information-rich near-infrared band in hyperspectral image, replacing the red band in the RGB image extracted from hyperspectral data to synthesize a pseudo-color image. Secondly, the pseudo-color image is matched with the true color image by histogram matching to maintain the color fidelity of the original image. Then, ACE-HF framework then uses an improved “decomposition-composition” ACE algorithm to enhance the image details of the stained area while reducing the computational complexity. Finally, homomorphic filtering is introduced to further separate noise information and improve the image quality of the lacquer Coffin painting. The experimental results demonstrate that the proposed ACE-HF framework method can extract and enhance the pattern information in the polluted areas of lacquer Coffin painting and has the potential to be applied to other contaminated cultural heritage images for implicit information extraction and blurred cultural heritage images for detail enhancement.