Design And Test Of Target Spraying System Based On Machine Vision

China is the largest country in vegetable production and consumption in the world.At present,vegetable production has become the second largest crop in China’s planting industry after grain.However,pests and diseases during vegetable production not only affect the appearance of vegetables,but also affect their growth,yield and quality.At present,chemical pesticides are still mostly used in the prevention and control of vegetable pests and diseases,and the application technology is outdated.Continuous spraying is used in the case of large vegetable row spacing and plant spacing.Deposited and lost to the air and soil,the utilization rate of pesticides is low and the waste is serious,resulting in environmental pollution and excessive foliar pesticide residues.In view of the above problems,this paper takes the field cabbage as the operation object,and designs a target spraying system based on machine vision according to the prominent color characteristics of field cabbage,which changes the continuous spraying method into intermittent spraying,and the nozzle spacing can be automatically based on the actual crop planting row spacing.Adjust the line to achieve precise application,effectively reduce pesticide waste and avoid environmental pollution.The following main research results were obtained:(1)The walking chassis of the target spraying system is designed.The walking chassis of the target spraying system includes the frame,the walking wheel,the steering mechanism and the spraying working parts.The spraying working parts are composed of three parts: the nozzle spacing adjustment mechanism,the nozzle height adjustment mechanism and the spray pipeline.The walking chassis adopts electric wheels and automatic steering trapezoidal mechanism.Combined with the control system,it can realize remote control and automatic walking,so as to achieve the goal of separation of people and vehicles and separation of people and medicine;the nozzle spacing adjustment mechanism uses the parallelogram frame characteristics of the truss rod structure to realize the spray nozzle spray.The continuous adjustment of the spacing can achieve the purpose of spraying on the row;the height adjustment mechanism of the nozzle realizes the adjustment of the working height of the nozzle through the jack structure and the telescopic spray rod,so as to meet the application requirements of crops of different heights;the spray pipe liquid is added by the diaphragm pump.The pressure is then controlled by a solenoid valve to spray.(2)The hardware circuit and software program of the target spray control system are developed.Taking STC89C52 microcontroller as the control core,the Bluetooth module is used to establish wireless communication between the mobile device and the sprayer,and the sprayer communication interface is designed on the smartphone,which can send commands to control the walking of the sprayer and adjust the spraying parameters,and can also display the sprayer Vehicle speed and medicine tank liquid volume information;PCF8591 module is used to output analog signal to realize the adjustment of the speed of the walking wheel,and the AD analog input channel combined with the sensor realizes the collection of the walking speed and the medicine tank liquid volume information;the drive circuit realizes the microprocessor’s control of the motor,Control of actuators such as electric push rod,diaphragm pump,solenoid valve,electric wheel,etc.;use Open MV vision module to realize crop target identification,and control the spray device to automatically apply medicine to the target according to the target orientation and other information.(3)The image recognition and segmentation algorithm of green crops is studied.The image segmentation effects of different color factors and threshold segmentation methods were analyzed,and an image segmentation method of Datian cabbage based on Gaussian bilateral filtering and ultra-green algorithm was proposed,which could achieve accurate segmentation of cabbage images under different lighting conditions,and based on edge The detection extracts the outline and center point features of Chinese cabbage.(4)The performance of the target spray system was tested.In order to test the performance of the target spray system,the response time of the control system,the line error,the uniformity of droplet distribution and the actual spray effect in the field were tested.The results show that: the response time of the control system is 344ms;the row deviations of the crop rows of 45,55,and 65 cm are 7.6,8.2,and 7.2mm,respectively;The best working parameters determined by the second rotation orthogonal test are the walking speed of 0.39m/s,the spray flow rate of 340.76 m L/min,the spray distance of 42.22 cm,and the minimum deposition variation coefficient of 13.11%;,60 cm drug saving rate were6.13%,25.80%,37.04%.