An Android Application-based Surface Oust Classifier

Surface dust can be a source carrier of viruses, bacteria and air pollution which entail common health issues such as asthma attacks, chest tightness, wheezing difficulty in breathing. Visually perceived, cleanliness is one measure indoor quality the subjective assessment cleaning quality. The aim this work to use pattern recognition mediated through mobile application analyse classify households, order obtain useful information about sources for selection appropriate countermeasures view improve better manage surfaces. type categorization are pollen, rock ash. This paper also explores concept transfer learning techniques adopts it small particle classification using CNN models by developing surface android smartphones. behaviour InceptionV2, InceptionV3, ResNetV2, MobileNetV2 MobileNetV3 feature extractors were analysed based on their accuracy, precision, recall Fl-Scoreperformance metrics. Results show that model best suited extractor rapid prediction with an accuracy up 92% low-size storage only 30 megabytes. Additional Keywords: Android; Classification; Transfer Learning; Smartphone; dust.