To solve the problems of traditional garbage classification relying on manual work, low recognition accuracy, and unhygienic contact-based disposal, an inductive intelligent garbage recycling system based on a micro-controller is designed. The system takes STM32F103ZET6 as the core, integrates speech recognition, visual recognition, and infrared induction modules, and adopts a dual recognition fusion strategy of “speech priority + visual verification” to realize con-tactless induction lid opening, accurate garbage classification, and remote status monitoring functions. Speech recognition completes keyword matching through the LD3320 module, visual recognition realizes image feature extraction and classification based on the BP neural network, and the ESP8266 module is responsible for uploading device status to the Cloud platform. It is applicable to multiple scenarios such as households and communities, effectively improving the efficiency and convenience of garbage classification, and has good practical value.