RECONMATIC partners from Czech Technical University in Prague, Vitek Stanislav, Zbíral Tomáš, and Nežerka Václav, have published a new study, "Using Hyperspectral Imaging to Identify Optimal Narrowband Filter Parameters for Construction and Demolition Waste Classification". The research focuses on optimising hyperspectral imaging (HSI) for construction and demolition waste (CDW) classification.
HSI is widely used for material analysis, but its high data requirements make real-time waste sorting impractical. Instead, the study explores using selected narrowband filters with standard cameras to achieve similar accuracy with reduced computational effort. By analyzing CDW materials with a hyperspectral camera and applying a multilayer perceptron classifier, the researchers identified two key wavelength ranges—650-750 nm and 850-1000 nm—as optimal for classification.
This discovery highlights the potential of near-infrared imaging to improve CDW sorting efficiency, supporting sustainable construction practices.