Maximum Contrast Imaging Using Chromatic Lighting and Fuzzy Logic: Analyses of Old Handwritings

Machine vision has been applied to many industrial areas. It has also become popular in heritage science. This study proposes maximum contrast imaging to determine the appropriate light color and intensity to increase the discriminating ability for image analyses. The contrast of an image was evaluated in accordance to various digital focus indices (DFIs), and their formulations were thus investigated. Optimal conditions of focus, light color, and intensity were determined based on the maxima of the DFIs. Fuzzy logic was applied to verify the determined owing to the different DFI maxima. The proposed method was implemented subject to the assumption of an arbitrary number of light colors. In the experiment, the DFIs were constructed using C/C++ libraries compatible with OpenCV and the decision methods were verified using an automated optical inspection (AOI) machine. Test targets were old handwritings in a xylograph book, and optimal conditions were determined in the case of a multi-color light source. The handwritings were initially illegible but became readable using maximum contrast images. This study showed that maximum contrast imaging will be useful in industrial applications, such as automatic optical inspection, and in the analyses of historical objects.