A Comprehensive Review on the Identification and Classification of Leaf Diseases Model by their Various Stages

Abstract

An important precondition for economic progress is dependent on agricultural cultivation, regardless of wherever in the world. Many problems are related to the divergence in agricultural crop output that estimates around the globe such asoveruse of inorganic pesticides and fertilizers, the presence of hazardous substances in water supplies, irregular rainwater dispersion, varying fertility of the soil, and other factors are among them. Aside from these concerns, amongst the most prevalent concerns throughout the world is the loss of a large portion of agricultural yields due to plant leaf infections/disease. After giving natural resources efficiently to the farms, the existence of fungal or bacterial pathogens in the developed plant's leaf reduces a large portion of the yield. As a result, researchers are concentrating on efficient automation methods for detecting leaf diseases within plants. Farm owners are concerned about the existence of numerous diseases in plants. Fungal pathogens are the biggest challenge to local farmers because they cause chaos in the farming industry. Primary prevention of the kind of plant diseases existing is required to give efficient procedures for diagnosis and prevention of damage.Whereas many articles mostly on recognition of plant leaf disease using various tools and techniques have already been issued in past decades, it seems to be no comprehensive examination of such articles to portray different modules of disease detection such as "Preprocessing" to eliminate the noises in the image, "Segmentation" for segmenting the affected region, "Feature Extraction" to extract the statistical features, "Feature Selection" for selecting the optimal features, and "Classification" for classifying particular leaf diseases altogether. As a result, we suggest this unique survey paper that includes all of the modules that make up the automated "Leaf Disease Detection" (LDD) software/tool. Throughout all stages of the LDD system, several concerns currently remain to be solved, according to this systematic review. The benefits and drawbacks of the reviewed methodology are discussed, along with prospects. The results of this investigation show that automated detection and classification approaches for LDD are still in their formative stages. As a result, innovative completely automated technologies are required for the detection and classification of plant leaf diseases.