Prediction Of Chlorophyll Content In Leaves Of Cucumis Based On Image Processing

With the breakthrough of 5g technology in China,Internet of things technology has become a hot vane.In agriculture,intelligent agriculture will be a new development trend in the world,and it will lead agriculture into a new digital and information age.It will also be the key direction of agricultural development in the world in the 21st century.The core idea of intelligent agriculture is to form a huge data network from the bottom sensor data,vegetation growth parameters and other aspects of agriculture.So that professional engineers and agricultural technicians can use data to guide agricultural irrigation and fertilization.Using advanced neural network and image processing technology to identify,analyze and measure the chlorophyll content in the current plant environment is of great significance and benefit to us.We can help us quantify the current rate of photosynthesis by measuring the content and conversion rate of the main chlorophyll species in the current plant.We can also help us know the current physiological health status and the lack of fertilizer in the plant through the content of chlorophyll in the plant.Only by mastering the growth and nutrition of current crops accurately,can we reasonably input fertilizer to prevent it from wasting and seriously polluting the surrounding environment of plants.In this paper,based on the previous research results,the image preprocessing process is improved,including normalized illumination removal,median filtering method,image segmentation method using Otsu segmentation.method.This paper further proposes a method of image segmentation using Dazu method after removing the illumination of image based on normalized preprocessing,and achieves a better preprocessing effect.In terms of color model,RGB and HSI model[39]are used.After mathematical transformation of six channels,R,G,B,h,s and I,four parameters(R,R-B,h,R1)with the best correlation with chlorophyll content are extracted.Finally,SPSS software is used for correlation analysis,and good results are achieved.Through the cultivation of melon plants under the same conditions,and the use of intelligent Huawei p20pro mobile camera to collect photos of melon leaves,the light conditions are natural light in the greenhouse,and the use of leaf chlorophyll tester to determine the SPAD value of chlorophyll as validation information.After collecting the SPAD of the leaf image of papaya,the regression equation and neural network model were drawn.Therefore,this study tries to find a good prediction model and a group of effective representation information by measuring the actual SPAD value of the blade and fitting the predicted SPAD value calculated by image technology and neural network technology.