2009 年 47 巻 3 号 p. 145-156
This paper highlighted a new methodology for quality control of an Affective Bio-greening Material (ABGM). ABGM is a bio-greening material which is produced in non-uniform controlled environment to generate the dynamic appearance for satisfying the different consumer. In this methodology, the consumer evaluated easily the material affectivity using its appearance in the questionnaires. Subsequently, this appearance was detected in production using textural features extracted from grey level co-occurrences texture analysis and water content. Hence, quality control was defined as the function of irreversible change of textural appearance which was stimulated by water content. Artificial Neural Network (ANN) was used to develop an intelligent watchdog model. The model was defined that material affectivity can be controlled from textural images and water content in a given local temperature and relative humidity. The model was demonstrated via a case study of Sunagoke moss greening (Rhacomitrium japonicum). The trained ANN model generated satisfied correlation between measured and predicted values and minimum inspection error. The research results concluded the possibility to control the affectivity of a bio-greening material in a non-uniform environment based plant factory production using an intelligent watchdog model.
