Visualization of Faded Handwriting Using Nondestructive Hyperspectral Imaging

Over time, the handwriting in paper documents fades due to the effects of the environment and materials, resulting in reduced readability. Hyperspectral imaging technology, however, can provide rich data information by collecting images at multiple wavelengths, which holds significant importance for cultural heritage protection. Therefore, this study aims to investigate the application of hyperspectral imaging technology in revealing faded handwriting in paper documents. In this study, samples with different inks were prepared for an accelerated aging process using a UV aging chamber to simulate the natural fading of handwriting, thereby obtaining samples with faded ink. Subsequently, a hyperspectral imaging instrument was used to collect and analyze data from the samples. The findings indicate that hyperspectral imaging technology can effectively restore the morphology of faded handwriting and significantly enhance handwriting recognition through preprocessing, minimum noise fraction rotation, and band extraction. Additional tests comparing ink and paper colors before and after scanning further confirm that the hyperspectral imager has a negligible effect on ink and paper colors during the scanning process, and the experimental method is feasible. This study demonstrates the effectiveness and potential application of hyperspectral imaging as a noninvasive means of recovering faded handwriting in paper documents.