Design of Intelligent Garbage Classification System in Shanghai

A design scheme of intelligent garbage classification system in Shanghai was presented which was applied to front-end processing of garbage classification in Shanghai. The system used speech technology to recognize garbage words automatically. At first, the system established a speech corpus of garbage vocabulary of five age groups which contain children, teenagers, youths, middle aged people and old people, including 15 men and 15 women per age group. Then characteristic parameters pattern base was set up by extracting Mel Frequency Cepstral Coefficients (MFCC) of speech samples via the use of MATLAB programming. Finally, based on Dynamic Time Warping (DTW), it can recognize categories of garbage automatically, including recyclable waste, hazardous waste, household food waste, residual waste and specially disposed waste, such as pet corpses and pet feces. Based on the results, the accuracy of the speech recognition rate for specific person and non-specific person reaches 100% and 98.78%.