Research On Sunflower Production Measurement Method Based On Machine Vision

Sunflower is a one-year-old herbaceous plant in the genus Asteraceae and is known as one of the four major oil crops in the world with important economic value.Precision agriculture is the product of the application of modern information counting in agricultural production,and the acquisition of output is the key to the implementation of precision agriculture.The capacity of sunflower has the characteristics of spatial change in the farmland,which also effectively reflects the growth and management of agricultural products.Many researchers at home and abroad have proposed a variety of production monitoring methods,but due to the influence and limitation of various factors,the accuracy of production measurement and the versatility of the model cannot fully meet the requirements of users.In recent years,with the further development of network technology,bioengineering technology and environmental engineering technology,machine vision technology has been widely used in the field of agricultural engineering due to its non-contact,accuracy,speed,intelligence and other characteristics.Therefore,this paper proposes to use image processing to monitor sunflower yield.This subject takes the "SH361" hybrid as the research object,and conducts the following researches:(1)On the basis of studying the performance parameters of each hardware,this article selects the CCD camera,lens,Online pictures and videos use collection cards,and ring-shaped LED light source is used for lighting and other equipment,and builds a hardware platform for the machine vision system to collect sunflower sunflower image.The basic equipment of the system includes components such as image acquisition equipment,experimental operation area,CCD,lens and computer.(2)Formulate two image processing schemes.The preprocessing of the image includes grayscale conversion,image filtering and denoising,and binarization.Image filtering module is the most effective way to evaluate the filtering of multiple images.The maximum inter-class variance method is used to segment the binarized image,and the mathematical morphological processing is used to "open" the image and the area-based threshold method is used to effectively eliminate noise and obtain the idealtarget area.(3)Use the eight-neighborhood algorithm to calculate the area information of the largest connected area marked by the two schemes.The area information of a single grain is obtained by the area information and the number of grains obtained by manual counting,and then the information of the area of a single grain is determined by multiple sets of experiments.Next,the area of the sunflower plate and the area of the single grain are used to calculate the number of grains in a single sunflower plate.The method of taking one thousand kernels to obtain the output of a single plate and the total output.In this paper,through multiple experimental tests on the system,the experimental data is obtained,and the cause of the error is analyzed,and then the method of eliminating the error is summarized.The results of the study indicate that the production system has achieved relatively satisfactory results to a certain extent,but this method will be unstable when measuring crops that are not properly bred and cultivated.