Digital Spectrophotometry Aplied to Solid Waste Processing

Solid waste sorting is usually faced adopting selection strategies able to mimic human behavior, that is defining and implementing procedures based on the detection of object properties and attributes linked to parameters usually detectable by human senses. In this perspective pictorial attributes and morphological and morphometrical characters are the most widely used. Independently from the intrinsic difficulties to develop affordable and reliable sorting architectures such parameters usually well satisfy recognition problems except in some cases where i) a more strict classification is required and when ii) same detectable attributes belong to different materials presenting different physical characteristics and behavior for the further recycling. In this work are outlined, and critically evaluated, the problems arising when a full imaging digital based spectrophotometric approach is applied in the field of solid waste recycling, with particular reference to cullets sorting, both to achieve an higher degree of discrimination in terms of color and in order to perform the recognition of poisoning "ceramic glass" fragments.