MAJOR CHALLENGES ON USING MACHINE LEARNING AND DEEP LEARNING TECHNIQUES TO DETECT LEAF DISEASES IN ASIAN COUNTRIES
YOGESH KUMAR RATHORE *
Department of Information Technology, National Institute of Technology, Raipur (CG), 492001, India.
REKH RAM JANGHEL
Department of Information Technology, National Institute of Technology, Raipur (CG), 492001, India.
*Author to whom correspondence should be addressed.
Abstract
The use of computers in agriculture plays an very important role to increase the productivity of any country. The Asian counties having a rich verity of fruits, vegetables, and crops. In this paper, we are sharing a survey on different leaf diseases found in crops, fruits, and vegetables of Asian countries and also focus on how the machine learning played an important role in the classification of such diseases in recent decades. The paper is arranged into four major sections, in first section most common leaf diseases found in Asian country’s crops and vegetables are discussed along with the various feature extraction techniques used for detection of such diseases. Then in next section, recent machine learning and deep learning based techniques which have been used for recognition and classification of plant diseases are focused, along with challenges and limitation of each technique. In the third section, we have evaluated all the techniques discussed in the previous section on the basis of different parameters like accuracy, complexity of algorithm, type of model and used dataset in order to find out the gaps in the existing system and to highlight the paths for new researchers in the same area. Finally, in the last section, some possible solution approaches are discussed for improvement in performance of the system.
Keywords: Fungal disease, bacterial disease, machine learning techniques, machine learning techniquesconvolutional neural networks, ROI extraction, leaf features, deep learning methods