Research On Chlorophyll Content Based On Tomato Leaf Image

The application of Internet of Things technology has become more and more extensive,especially in agriculture.Precision agriculture has become an important direction of agricultural development today.The essence of precision agriculture is to detect the process of crop planting in real time and accurately.Then analyze the internal information in the system,and compensate from the acquired information in a targeted manner to achieve the desired optimal effect.Based on the current development situation,this paper applies image processing technology to the prediction of tomato leaf chlorophyll content and soil nitrogen content.First,the tomato leaf image is preprocessed,and an OTSU segmentation algorithm based on improved weighted particle swarm optimization algorithm is proposed.This method improves the efficiency of image segmentation and extracts the characteristic values of tomato leaves after image processing.The SPAD value and the extracted characteristic values are used to establish a prediction model using linear regression,BP neural network,and RBF neural network algorithms,and the root mean square is used.The error RMSE and the value of the coefficient of determination(R2)were used to analyze and evaluate the three models.The model with the coefficient of determination close to 1 and the root mean square error was selected,and the chlorophyll content and nitrogen content of tomato leaves were predicted.The following work was completed:(1)Cultivate tomato samples,collect tomato sample leaf images,relative chlorophyll content SPAD value,tomato leaf nitrogen content and soil total nitrogen content,perform image preprocessing operations on tomato sample leaves,and propose a particle swarm optimization with improved weights Algorithm.Experimental results show that the algorithm improves the effectiveness of image segmentation processing,and finally an image containing only the leaves of tomato samples is obtained.(2)Select the RGB color space as the research object,extract the color feature parameters related to the chlorophyll content of tomato leaves,and select the feature value parameters that have a greater relationship with the relative chlorophyll content SPAD value according to the Pearson correlation coefficient(3)According to the selected eigenvalue parameters,linear regression algorithm,BP neural network and RBF neural network are used to establish mathematical models between relative chlorophyll content SPAD value,leaf nitrogen content,soil nitrogen content and eigenvalues,and calculate Based on the values of the root mean square error RMSE and the coefficient of determination,a model with a coefficient of determination close to 1 and a smaller root mean square error was selected,and the model was used as a prediction model for tomato leaf chlorophyll content and a prediction model for soil nitrogen content.This paper combines image processing technology with regression modeling analysis.Through image processing technology,the tomato leaf graphics are processed,the tomato leaf color feature values are extracted,and the mathematical model between the color feature value and the relative chlorophyll content is established.Prediction of chlorophyll content,rapid and non-destructive testing of chlorophyll content of tomato leaves and prediction of soil nitrogen content.