Research On Inspection Of Lettuce Nitrogen Nutrition Based On Computer Vision

Lettuce leafy vegetables are indispensable in people’s daily lives, providing the body with important nutrients.In recent years, in order to increase production, excessive fertilization of leafy vegetables are common occurrences.As the main factor for affecting crop quality and yield,the amount of nitrogen fertilizer should be decided under accurate know the level of demand for the crop.In this paper an accurate、non-destructive、real-time plant nitrogen monitoring method has been research. To achieve rapid、real-time crop nitrogen diagnosis in agricultural production is the key to reasonable application of nitrogen fertilizer.In this paper,loose leaf varieties Italian bolting lettuce for the study,discuss about image acquisition problems、the color parameter selection problems、modeling problems in lettuce leaves-Nitrogen Content based on computer vision designed to lay the foundation for the development of the crop nitrogen monitoring platform based on computer vision.The main results are as follows:(1)Through the different levels of nitrogen treatments,Research lettuce growth characteristics under different nitrogen Levels,analysis lettuce-leaf stage nitrogen content, SPAD, leaf differentiation under different nitrogen Levels, Lay the foundation for analysis of lettuce leaf nitrogen content and color parameters.(2)Research lettuce canopy image acquisition methods in greenhouse environment,Contrast image information collected by the two kinds of image acquisition hardware Canon550D and wave-particle HD ball machine,Come to the conclusion that images information collected by the Canon camera accuracy better than wave-particle HD ball machine.(3)Analysis correlation between nitrogen content and color parameters of the Lettuce canopy image、calculate the value of the color parameters. Research derived:, In the RGB color model,R、G、B、R/G、G/R、g、r values were highly significant(4)positive correlation with the nitrogen content.G-R value is highly significant negative correlation with the nitrogen content.The correlation between nitrogen-containing and chromaticity coordinate value of r、g is lower than the R、G value.In HSB color system, H、B value is significantly correlation with nitrogen content.B value is a significant positive correlation with the nitrogen content.H value is a significant negative correlation with the nitrogen content.(5)According to the results of correlation analysis between lettuce leaf color parameters and nitrogen contents,Select significant correlation of12key color parameters R、G、B、H、B(HSB)、r、g、G、R/G, G-R、G/R,develop Lettuce leaf nitrogen content monitor model.Respectively,using principal component analysis method and multiple regression analysis develop Lettuce leaf nitrogen content monitor model and validate the model.