Automated Trash Separation

CSL | Computational Sciences Laboratory


Automated Trash Separation Using Pattern Recognition and Image Processing

Automated Trash Separation involves the activities associated with management of waste articles available in trash by segregating them into various classes and subjecting them for further analysis. Separation of waste components is an important step in the handling and recycling of wastes. This involves segregating them to various classes like paper, plastic, cups, plastic bottles, glass bottles etc. Till date, this procedure is being done manually which can be automated to facilitate recycling and effective classification.
Pattern Recognition involves computational technique used to find patterns and develop classification schemes for data in very large data sets. Here we try to extract specific features which can uniquely identify a particular object and this feature is fine tuned to obtain finer details of the object to provide the sub-classification of within large classes of data sets.
With the help of several image processing techniques coupled with pattern recognition methodologies, we can bring out software to facilitate automated Trash Separation.
This involves discipline of Machine Learning, having to deal with properly maintained large data sets, and having to derive rules from learning history which involves Software Engineering Methodologies.