Grey image recognition-based mold growth assessment on the surface of typical building materials responding to dynamic thermal conditions

Abstract

Mold growth on building materials poses a threat to both the building structural integrity and occupants’ well-being. However, it is generally studied under suitable conditions in laboratory; the assessment is based on visual inspection but lacks an objective criterion. This research explored the effects of simulated dynamic thermal conditions on mold growth on three typical building materials artificially contaminated with Aspergillus niger spores. Test specimens were assessed based on a developed digital image-based method, where image segmentation, processing, and greyscale recognition via the OpenCV visual library were introduced. The results showed that the high temperature-high humidity condition in a 24-h cyclic change facilitated mold growth on the surfaces of three materials, especially for gypsum board, with an identified area proportion of 1.13% on the 80th day. This was consistent with the changes of the counted number of mold colonies, and no significant differences were found among the gypsum board, latex paint, and wallpaper. The growth extents of mold spores were objectively evaluated by the mean greyscale values; the values decreased gradually with time, and the decrements were different compared high-temperature to low-temperature conditions. The mold growth models were developed, where the area proportion of mold growth was linearly related to the counted colonies and greyscale values under different material surfaces. This novel grey image recognition-based method provides an accurate means of evaluating mold growth abilities and extents, overcoming the inaccuracy of visual observation. The findings have significant implications for visual inspection, mold prediction, and building management.