ISSN 2348-1196 (print) International Journal of Computer Science and Information Technology Research ISSN 2348-120X (online) Vol. 9, Issue 1, pp: (127-135), Month: January - March 2021, Available at: www.researchpublish.com
Apple Leaves Disease Detection Using Machine Learning Approach Kulbir Kaur Sandhu Assistant Professor, Department of Computer Science Baba Farid College, Bathinda kulbirkaursandhu01@gmail.com
Abstract: Agriculture plays an indispensable role in the development of the country especially in the growing country like India where most of the people’s revenue is generated from agriculture. Disease affected crops leads to the loss of crop productivity. Therefore, leaf disease prediction in apple cultivation is of considerable importance to overcome these problems. The proposed work intends to predict different disease in apple leaf like apple scab and marssonina using different algorithms like K nearest neighbor(KNN),support vector machine(SVM), classification decision tree, regression decision tree and Naïve Bayes. From the simulation result, it can be concluded that KNN performs better as compared to other algorithms in terms of accuracy of disease prediction. Keywords: SVM, KNN, Apple disease, Marssonina, Naïve Bayes, inverse, Apple scab.
1. INTRODUCTION India is ranked number two in the production of fruits. Sixty percent of the population depends on agriculture for employment. The methodology used for fruit illness recognition is perception through naked eyes by specialists. In some countries, such specialists are costly and tedious because of they are not accessible. Damage can also be seen in areas such as stems, leaves, and branches of the tree . Apple rot,apple scraband apple blotch are some of the most common diseases of apple fruits. Apples with gray or brown corky blotch have apple scab diseases . If the surface of the apple have black specks that may be covered by a red halo or the fruit is faintlyhallow, circular brown then, it is referred to as apple rot type diseases. Apple blotch type diseases have surface with dark,irregular or lobed edges .The diagnosis of apple diseases are mostly performed by an expert, on site by inspecting the leaves and fruits visually. In more complicated or new cases,analysis in laboratory can be done. However, it needs experts that demand specialized training and this approach increases the overall cost. The need for experts limit scale. Since experts are often specialized in a few types of disorders the effectiveness may reduce. There are mainly two type of images and these are Analog and Digital image. Processing of image means to turn an image into its digital form and then different kind of operations will be performed on it to get some kind of results with enhanced type of same image. These results will give us the useful information which can be extracted from it and can be very helpful. There are many type algorithms which can be implemented to perform a particular type of operations or task. Sometimes when some diseases are not visible to naked eye but actually they are present, then it is difficult to detect it with the naked eye. And when it is visible it will be too late to detect disease and can’t help anymore. Earlier, this disease was detected with the microscopes but it was very difficult process to observe each and every plant and leaf. So now remote sensing techniques are developed and used in computer science. They can detect the disease from the multilevel and hyper spectral images of plants which can be digitally captured.
Page | 127 Research Publish Journals