Application of Feature Library Modeling and Morphology Recognition Technique in Reverse Design on Sorting Process of Recyclable Wastes

The sorting and recycling of recyclable wastes in urban life is important to implement the national "dual-carbon" strategy and enhance the quality of residents' living environment. In this paper, a comprehensive technical solution for sample measurement, curved surface reconstruction, and model feature library construction of recyclable waste samples is proposed based on the 3D feature modeling and morphology recognition technique. The process in this solution includes discrete coordinates measurement of sample points in model surface, segmentation feature reconstruction of curved surface, regularization treatment of modeling error in local curved surface, integrated reconstruction of curved surface features, and feature library modeling and morphology recognition. Results from the example verification demonstrate that the application of feature library modeling and morphology recognition technique in reverse design on the sorting process of recyclable wastes would effectively improve the recognition accuracy of wastes morphology. Furthermore, the reverse design technique is applied to mitigate distortion in feature sections during feature modeling, as well as reduce ambiguity in morphology recognition. The modeling results based on feature library provide complete data models, including geometric information. The solution has a broad prospect of application in sorting and selection processes of the similar types of urban wastes and the technique is an important foundation and supporting one for additive manufacturing (3D printing) applications.