Lossless Detecting Research Of Water Quantity In The Leaves Based On Image Processing Technology

Along with the continuously development of computer technology, image manipulation technology is not only widely used in every walk of life but also play an important role in the crop growth in the greenhouse and scatheless inspection. The use of image processing technology for agriculture inspection has become an important matter in agricultural automatization, apply image manipulation technology to the image you get and transform it or analysis it , in order to fulfill people's need of agricultural produce and other factual purpose, it's an inevitable trend.Irrigative automatization is one crucial control process in the automatizatic agricultural produce.Vegetable is one crop which has a great demand of water. Compared with other crops, vegetables have more consuming response to water. Irrigation is frequent during the growth period of vegetables, the output depends on whether we irrigation in time or not. If we crop vegetables in the greenhouse or other kinds of dependency, it will not only be propitious to increase production and saving water, but also improve the quality of vegetables.Base on domestic agricultural production situation and the existing literature domestic and overseas, the main purpose of this text is to research the image manipulation and the arithmetic of it. Unlike conventional method of leaves water quantity inspection,this text bring forward a new method to test the water quantity in the leaves of forcing cucumber by using image manipulation technology. Using visible light with different wavelength to irradiate the leaves from the background in order to gather the image, then confirm the water quantity from the image gathered, consequently study the water condition of the plant. Replace conventional method of leaves water quantity inspection by measuring the water quantity of the living body's leaves, achieve the lossless detect of the plant's water quantity.This has changed the crop in the past studying aspect of middle difficult to determine the amounts of the components of a substance-rization, italso can realize information collect automation and intellectualized together. Throughout the leaves image outward appearance characteristic to carries out detecting on water quantity of the crop leaves is one kind of rapidly simple and direct , practical effective method. We can making use of this method to irrigation the hothouse plants according the stage and growth condition in a more rational way, also it provides a basis to the extensive use in helping advance career to look forward to saving water in the crop using the image manipulation technology.The main content of thesis is arranged for as follows:Firstly, introduced the significance of checking vegetable water quantity briefly and the main detecting method at present, then introduced some basic concept of image manipulation technology, including the development, characteristic, application field, development of the days to come subsequently and the application of checking in the agriculture of it.Secondly, according to the process design, systemically expatiate the basic theory of image manipulation technology, and structured a loseless detecting system of water quantity of the cucumber leaves by means of image analysing.First of all structured the image acquisition system of leaves, including the hardware equipment and the basic digitized principle of image, gained the image of cucumber leaves, laid down basis for the follow-up image pretreatment and the whole design of the detecting system. Adopt the CCD digital camera to collect image, making the system have the characteristics of high resolving power, the operation is brief fast, the check-up measure is lossless not hurt the leaves and have and detect it in the real time. In the whole design of the amount of water examination system, introduced the system principle, the software and hardware design, combine the the amount of water examination system this text elaborate the research conception, the research method. The difficulty of the system lies on the select of background illuminance and method, because different colour have different character, also because the instability of the naturalness light causes flickering, it brought about degree of difficulty on distilling the imagecharacteristic, therefore the experiment has designed a set of experiment scheme specifically design main on the illuminant colour choice for the illuminator and background illuminant and illumination carrying out a lossless detecting system in the greenhouse.Since the computer image handles a technology damaging main difficult point of detecting lying is the background ray, the experiment design a series of experiment such as choosing an experiment by the fact that the contrast test, the experiment of visible light colour select , the most suitable number of LED and so on. Finally ascertained the illuminant condition and ascertained tne system composition ultimately.Thirdly, study the various arithmetics. In the image pretreatment, mostly studied the multifarious filter technology, including the median filter, the equal value filter, and analysis the characteristics and application of the filter technology. Subsequently studied the RGB and HIS color space and the like, later on studied the relationship with brightness. In the contrast with other arithmetics the tone soft grey could transform with RGB one another, the RGB can convert to HIS, we adopt the tone soft grey arithmetic for the simple and fast reason. We make use of the grey arithmetic to analyze the image characteristic of the leaf, distilled the characteristics of the main district that carried on two sides of the leaf. Whether the crop is short of water or not manifest whether it is damp off. If it is rich in water, the main nervation has great brightness contrast in vision; if it is damp off, the main nervation will become dark, the luminance contrast will not as great as that of the nervation rich in water. Therefore this experiment distill of the main nervation's characteristic is the most important contents that this text studied. The experiment distill 3 pixels width and 6 pixels length area and invert it into grey gradation form, then draw out the grey gradation changing a curve. Pick-up the characteristic that carried on two sides of the main nervation's districts of the leaf deal with the picture with MATLAB, at last anatomise the picture which has the greatest luminance contrast and select the most obvious grey gradation change photograph.Finally, through a series of experiment such as choosing an experiment by the fact that the contrast test, the experiment of visible light colour select, the most suitable number of LED, the critical point experiment and the analysis of the arithmetic selection, protract the connection curve of leaves'water quantity and image character under the optimum background light, and certain the cucumber water shortage threshold through the critical point experiment. This text uses Nonlinear Least Squares Method to make Log-Modified recursive models of the curve protracted through the experiment, and set up the Log-Modified model of the connection curve of leaves'water quantity and image character under the optimum background light, consequently gained the recursive equation. The decided correlation coefficient R2 is 0.9971, the prominence level p<0.00001, achieve the extremely prominent level and match the result well, the regression equation is: X2= (6.3569-6.7602*X1)-1.6802 .Consequently we can immediately judge the farm crop humidity plentiful lack the circumstance, whether need to be watered right away by making use of that curve, also we can know the water quantity at the moment by the recursive equation as long as shooting out the cucumber leaves image.