| The weed is a main influencing factor causing the declination of crop quality and yield,because it compete with crop for sunlight,space,moisture and nutrients.At present,application of chemical herbicides is the most widely used and effective weeding method,but it also cause may problems such as environment pollution and food safety issues.In recent years,as a safe,effective and environmentally friendly weeding method,mechanical weeding has become an important approach to solve the pollution from the chemical weeding.However,it is very difficult for the traditional mechanical weeding component to change its position.According to the changes in seedlings belt path that causing a critical seedlings injury during the weeding operation.And,the development of the mechanical weeding technology was restricted.Therefore,to realize seedlings avoidance control and reduce the seedlings injury rate,developing the rice seedlings belt identification and the operation path of weeding component differential regulation are the crucial issues.In order to avoid seedling and reduce seedling injury rate during the operation of mechanical weeding parts of rice,a seedling avoidance control system based on machine vision and fuzzy control was designed.Gray level were optimized by genetic algorithm operator and the seedling image gray scale,using the image projection to article levels within the region of the seedlings for feature extraction and image coordinates,based on the principles to improve the least squares fitting seedling get belt line,extraction of miao,from the center through the small hole imaging model transformation,seedling ground coordinates and weeding parts center and the seedling distance from the center line.The hydraulic rectifying system is controlled by using fuzzy control algorithm,and the automatic control of weeding parts is realized.The main research contents and conclusions are as follows:(1)The identification of rice plants in paddy field environment.In order to adapt to the identification and positioning of rice plants in paddy field,the color model was selected and the suitable color space for paddy field environment was determined.The image graying effects of different operators are compared,and the image graying processing based on genetic algorithm is proposed and realized.The image denoising technology was studied,and the effective identification of rice plants was realized by filtering,morphology and other techniques.(2)Extraction of center line of seedling belt.In order to realize the accurate identification and extraction of field thinning strips,the vertical projection method was used to identify and locate the seedlings divided into horizontal strips.This paper analyzes the commonly used method of extracting seedling strip line,and puts forward the method of fitting seedling strip line based on the improved least square method.The center line of seedling belt was obtained according to the seedling belt line.(3)Design of seedling avoidance control system.To make rice mechanical weeding unit operations can avoid seedling,to reduce the injury rate,studied the working principle of the control system of the avoidance,identified from the hardware composition and the type of the control system of the selection,design the software of the control system of the control program,avoid miao control based on fuzzy control algorithm is proposed,according to the lateral deviation error and the parameters of the lateral deviation error rate feature selection membership function,determine the fuzzy control rules,and established the simulation model of the fuzzy PID control algorithm.The simulation results show that the response time of the control system is about 0.2s.(4)Performance test of seedling avoidance control system.In order to verify the performance of the center line fitting of the seedling belt,the performance test of the image processing system was carried out.The test results show that the maximum error of the center line fitting of the seedling belt is no more than 2.25 cm and the average error is no more than 0.86 cm under the conditions of different transplanting days,different traveling speed and different seedling spacing,and the system has a high fitting accuracy.The system has high fitting precision.In order to verify the performance of seedling avoidance operation,the soil tank test of the control system was carried out.As the result of test shown,the order of the factors affecting the injury rate from serious to mild is the transplanting days,the plant spacing and the moving speed.The seedlings injury rate was2.27% under the optimal condition that the spacing was 18 cm,working speed was 1.2m/s and transplanting days are 10 d. |
