Multi-material classification of dry recyclables from municipal solid waste based on thermal imaging

•Thermal imaging based dry recyclable material identification from MSW is presented.•Iron, plastic, paper, aluminum, stainless steel and wood from MSW are successfully classified.•The classification accuracy range is found to be 85–96%.•The study was applied for the recovery of recyclable objects present in simulated MSW.•Potential exists for large-scale implementation by municipal bodies where source segregation is not followed. There has been a significant rise in municipal solid waste (MSW) generation in the last few decades due to rapid urbanization and industrialization. Due to the lack of source segregation practice, a need for automated segregation of recyclables from MSW exists in the developing countries. This paper reports a thermal imaging based system for classifying useful recyclables from simulated MSW sample. Experimental results have demonstrated the possibility to use thermal imaging technique for classification and a robotic system for sorting of recyclables in a single process step. The reported classification system yields an accuracy in the range of 85–96% and is comparable with the existing single-material recyclable classification techniques. We believe that the reported thermal imaging based system can emerge as a viable and inexpensive large-scale classification-cum-sorting technology in recycling plants for processing MSW in developing countries.