Category: Classification of segmented image using increased global contrast for paddy plant disease

This chapter presents the detection, diagnosis and identification of diseases in the paddy leaf. The suggested method can be followed at each step for the type of crop and its leaf. The most important task is extraction of the features of the leaf and doing further analysis. The classifier is used to set the tone… Continue reading Conclusion

Simulated output of classification, image processing and segmentation of the infected paddy plant leaf is shown in Figures 6.7–6.10.

Flow chart of the proposed method.

Using different digital devices like digital cameras, mobile phones and laptop, we can capture the image of the leaf. Generally, image can be greyscale or color. Six images are considered standard as per the dataset. Out of six, three are colored and three are greyscale images. Our first task is to check whether the image… Continue reading Image acquisition

In advanced machine learning, the classification is done by SVM. The SVMs aim to create a decision boundary using which we can identify boundary from two suitable output values. SVM is also used for classification purpose. It is a supervised algorithm with which we can the change the boundary position also. The points closer to… Continue reading Support vector machine (SVM)

We did image classification using classifiers. Based on the degree of disease in a leaf, a classifier is used for classification and classification depends on the features of the last step. In machine learning it is used to divide values into two groups: one of positive values and other of negative values, depending upon certain… Continue reading Classification of image

Using the methods as indicated earlier, the image is characterized and segmentation detects the diseased area of the leaf. By using below formula, we can calculate percentage of the infected area of the leaf. In the formula, DA is the diseased portion of the leaf area and LA is the total leaf area (Figure 6.6): IA=DALA×100.

The size reduction of the image data is done by feature extraction. The segmented region is very easy for doing study as it is filled with suitable colors and shapes. Most of the features are used to illustrate an image of paddy leaf.

K-​​​means cluster process is applied to the image to do cluster analysis. It is the best method to solve problems related to clustering of the best learning algorithms that solve the clustering problems. K-​​​means algorithm is used to detect the iterative K-​​​centers. The sum of distances of the data points is optimized from their centers.… Continue reading K-​​​Means clustering