Abstract proposed system we identify the disease on

Published by admin on

Abstract            Due to increasing demand in agriculture and Economy, farmers are need to increase their productivity of crops at the same time they also earn more profit in their cultivation. This is very difficult for small farmers because there are so many challenges for farmers like canopy, water management, weed control,weather conditions, field conditions and less rainfall. This may  be achieved by precision agriculture. In this paper we proposed a fuzzy inference system based classification for early disease identification, and we used image processing techniques as an automatic tool which will help to reduce the communication gap between the small farmers and agriculture officiers. In the proposed system we identify the disease on rice plant leaf images using L*a*b color feature based segmentation. The segmented images were classified into healthy, affected and highly affeteced classes with the help of Fuzzy Inference system.Keywords: Plant leaf disease classification, Fuzzy Inference System, L*a*b, Paddy DiseasesI INTRODUCTION            India is a land of small farmers, today our farmers facing so many challenges on their cultivation because of unconditional weather and high population, the food production need is high but the farm land is lesser than our needings.so farmers are having the pressure of increasing their yield with few acres,due to this they use chemical pesticides in huge amount on their crop this may cause very serious problem on people health .These problems could be overcome by “smart farming” .     In this work we used rice plant images as a input to calssify diseases because rice is a primary food source of   India. Rice may lose its quantity and quality when rice plant is attacked by different disease. Therefore, it is a top priority to find effective methods to reduce the level of their infestation in the paddy fields. Leaves are the best health indicator of a plant so we use the green intensity values and the number of brown spots as a parameters to identify the diseases on rice crop. II Literature Review     In Past few years the image processing techniques widely used for agriculture in various problems like disease classification,weed management, fruit garding, pest identification .    Xuebing Bai e.t al(2017)9 segmented a cucumber leaf spot from its complex background by using fuzzy clustering based neighborhood segmentation with high accuracy.    K. Muthukannan and Dr. P. latha classified a unhealthy and healthy portion of a tomato leaves using fuzzy inference system with 95% of accuracy      R. Pydipati et. al identified diseases on citrus fruits using color feature segmentation,they use various color features and CCM with the accuracy rate of 96.3%       Libo Liu, Guomin Zho,Extracted rice leaf disease using back probagation neural network model with high accuracy.   III PROPOSED METHODOLGY                In this work more than 100 images were used for classification these images were collected from internet dark spot diseased images and healthy leaf images were used . Those images had various size and formats that are resized and used for further process.1.Image Preprocessing                          The images were collected from the internet were resized into 300x 300 size uniformly to increase the performance of algorithm and reduce the complexity. In many agriculture research papers they suggested the above mentioned size.2.Color Feature Extraction: Color is perceived by humans as a combination of tristimuli R (red), G (green), and B (blue) which are usually called three primary colors. From R,G, B representation, we can derive other kinds of color representations (spaces) by using either linear or nonlinear transformations. Several color spaces, such as RGB, HSI,CIE ¸*u*v* are utilized in color image segmentation, but none of them can dominate the others for all kinds of color images. Selecting the best color space still is one of the difficulties in color image segmentation1. 2.1.1RGB: Red, green, and blue components can be represented by the brightness values of the scene             obtained through three separate fillters (red, green, and blue fillters) based on the following equations1:                         2.1.2HSV: HSV is an  another color space which represented as cylindrical geometry with hue, their angular dimension, starting at the red primary at 0°, passing through the green primary at 120° and the blue primary at 240°, and then wrapping back to red at 360°. 11                                           HSV                                                 L*a*b                                                    YCbCr    2.1.3 L*a*b: This color space is originally defined by CAE and specified by the International Commission on Illumination  In this color space, we have one channel is for Luminance (Lightness) and other two color channels are a and b known as chromaticity layers1.  The a* layer indicates where the color falls along the red green axis, and b* layer indicates where the color falls along the blue-yellow axis. a* negative values indicate green while positive values indicate magenta; and b* negative values indicate blue and positive values indicate yellow.112.1.4 Ycbcr    The YCbCr color space is widely used for digital video. In this format, luminance information is stored as a single component (Y), and chrominance information is stored as two color-difference components (Cb and Cr). Cb represents the difference between the blue component and a reference value. Cr represents the difference between the red component and a reference value.  In this work RGB Paddy leaf images were converted into the above referenced color spaces with the help of MATLAB16 from those color spaces L*a*b color space chosen for color thresholding.                                                                                           Fig1:Color spaces on MATLAB   3. Color Thresholding                On the selected L*a*b space the channel second (a) has red and green intensity values .so it is selected for segmenting brown spots because leaves are green dominant, spots may have a combination of red and green intensity. On color histogram of *a space minimum and maximum  value stetted for separating brown spots                                               Fig:2 Brown spot diseased image                                                          Fig:3 Segmented image4.Fuzzy Inference System  The general operation of the fuzzy system is performed based on the following process.on MATLAB FIS editor supported for building fuzzy modelFig 4:Fuzzy model-agri leaf spot1. Fuzzification   The process that allows converting a numeric value (or crisp value) into a fuzzy input is called fuzzification. There are two ways to do fuzzification:where there is no noise. Triangular membership function was used to fuzzify input and output functions             Fig 5:Membership values of Green intensity input                     Fig 6:Membership values of brown spots input  In this work the paddy leaf were fuzzified  into three categories 1.Healthy leaf  2.Affected leaf and  3.Highly affected leaf for this fuzzification green intensity values and  no of spots are used as decision parameter The .Mean value of healthy green leaf and affected  leaf were calculated. Brown spot labels of normal and highly affected leaf images alsoCounted.   2.Fuzzy Rules        Fuzzy If-Then Rules Fuzzy sets and fuzzy sets operations are the subjects and verbs of fuzzy logic 8. If-Then rule statements are used to formulate the conditional statements that comprise fuzzy logic.A single fuzzy If-Then rule assumes the form.                                                                    Fig7:Fuzzy rules for agri-leaf spot identification2. Defuzzification       Defuzzification is the reverse process of fuzzification.Mathematically, the defuzzification of a fuzzy set is theprocess of conversion of a fuzzy quantity into a crisp valuei.e. rounding off from its location in the unit hypercube to the nearest vertex. This may be necessary if we wish to output a number to the user.Centroid method is used for deffuzification                                                        Fig8:Membership values of  leaf outputIV RESULTS  We develop agri leaf spot fuzzy model(mamdani) with two input variables namely green pixel and brown spots  for input and output functions triangular membership function was used.  Green pixel intensity of healthy images will be high but affected leaf green intensity will be lesser because affected portions dominated by red intensity.                    Fig 9:Affected leaf Result                                                    Fig 10:Healthy leaf ResultAbove Result green pixel range is 116 and no of spots is 55.7 then the leaf  categorized as affected. and the crisp output value is 0.5.Fig 11:Results of Healthy and Affected leavesV CONCLUSION   The main contribution of this paper was proposed an approach fuzzy rule-based system using color features for theclassification disease affected plant leaf image region. The proposed method FIS using color features were clearly classified the region such as healthy, slightly affected disease portion and highly affected disease portion of the plant leaf image Classifying the unhealthy region is the main purpose of the proposed approach. The experimental results indicate the proposed approach can detect the leaf diseases region with little computational effort. The extension of this work will focus on developing methodology for better segmentation and classification using neuro-fuzzy system.REFERENCES

Categories: Management


I'm Iren!

Would you like to get a custom essay? How about receiving a customized one?

Check it out