Design Of Dry Battery Recycling System Based On Embedded Machine Vision

China is the largest country in the production and consumption of dry batteries,waste dry batteries are always unfriendly to the environment,and they are more recyclable resources and should be re-treated.However,the current recycling method simply sets the recycling box in the residential area,and the poor recycling effect depends entirely on people's environmental protection and resource protection conscious.In order to solve the problem of collecting used batteries more effectively in residential areas,this paper developed an intelligent rechargeable battery recycling system based on embedded and machine vision.The basic functions of the system are as follows: the user inputs the object to be recovered,the recylcling system captures its image,then to obtain size data of the shape after the image processing,combineing the quality data obtained from the weighing measurement to determine whether the object is a certain type of battery,If it is,then recycle and write the corresponding bonus points to user IC card through RFID.The overall structure of the recycling system is composed of the bottom hardware platform layer,the middle system software layer and the upper application software layer.Hardware platform is based on STM32F407 microcontroller for the control and computing center,in whose periphery seting the camera module to capture image information,the RFID module to read and write user's IC card,the LCD for human-computer interaction,weighing module to measure the quality of the object,the motor module to drive the construction platform turning back and forth to preform recycling operations,and expending the FLASH and SRAM to ensure that the system has enough storage and running memory.System software layer includes: hardware module drivers,file system and memory management service program,the migration of the UCOS-?.The application software is the realization of the product function under the support of the hardware platform and the system software,it mainly includes the task of the weighing measurement and the task of the battery identification by the machine vision.The method of battery identification is: a frame of image data is captured and processed in grayscale.According to the histogram of the grayscale image,the automatic contrast enhancement algorithm is used for the image enhancement.Then the target binary image is extracted by the threshold segmentation and smoothed and denoised by the mean filter.Finding the minimum bounding rectangle of the filtered target image,and calculating the length and width dimensions of this rectangle,which will be compared with the standard size range of the dry battery.In this paper,we also produced a sample of the recovery system and conducted a system function test.The test proved that this system can accurately identify various types of batteries,recycle them and give users bonus points.The embedded system have the advantages of low cost,low power consumption.Also,our product is real-time,in good stability,and easy to use,which is suitable for promotion in communities or other residential areas.