Article Preview
Top1. Introduction
Food and Agriculture Organization of the United Nations Organization has supposed that food production increase by 70% until 2050 to fulfill the necessity of food in the world. Modern technology nowadays has provided ample space to produce to meet the demand of more than 7 billion people. Agriculture is the foremost occupation in India. In the whole world, India has secured its second rank in agriculture. There is a high chance of food insecurity if production is not increased. It can only be enhanced or increased using the exact topmost technology. Several factors are responsible for food security threatening. Among these, crop disease is the foremost threat to food security (Dhakal & Shakya, 2018). In earlier times, trained experts were used to detect diseases in the plant through the naked eye or visual inspection. It was costly and inappropriate, as there was a lack of human intelligence also. Machine learning was used to solve this problem in which CNN is its part of detecting the disease by preprocessing infected plants leaf and fitting into a neural network. According to Karol et al. (2019), by using image preprocessing, the first implementation of plant disease detection was performed by SHEN WEIGHEH WUYACHUN CHEN ZHANLIANG and WI HANGDA in their paper. According to Wallelign et al. (2018), they state that machine learning methods such as artificial neural network (ANN) decision trees, K-means, K nearest neighbors, and support vector machines (SVM) are used and applied in agriculture research. like machine learning, deep learning also contains supervised, unsupervised, semi-supervised, and reinforcement learning methods or models. Among the supervised models are applied in agriculture to perform different tasks such as image segmentation, image classification, and object detection. The production quality and quantity of tomato crops are affected by the wide-scale presence of disease. A study was completed by (Purushothaman et al., 2018) to counter the problem; images of tomato leaves were derived from plant village datasets as input to two deep learning-based architectures named Alex Net and VGG 16 Net. (Sumbwanyambe & Sibiya, 2019) states that grey leaf spot by maize disease is caused by Cercospora maydis fungus. It is supposed to be a serious threat to maize production in large areas of the US corn belt and Africa. A smartphone camera model was developed to recognize three different maize leaf disease types out of the healthy leaf. The diseases were northern corn leaf blight, common rust, and grey leaf spot. Common rust maize diseases caused by the Puccinia sorghi pathogen are favored by coo temperature (16-23 degrees Celcius) and high radiative humanity (100%). Moreover, spots are found on both upper and lower leaf surfaces. The present model can identify the following diseases in potato crops, tomato leaves, and pepper bells. These are bacterial spots in pepper bell; early blight and late blight in potato; bacterial spot, late blight, early blight, leaf mold, septoria leaf spot, spider mites two-spotted spider mite, target spot, tomato yellow leaf curl virus, and tomato mosiac virus in tomato plants. So for detecting the plant disease, a CNN model is developed. The aim and goal of our research is to solve the problem of detection and prevent diseases of agricultural crops. India is a land where agriculture is the main occupation of people. Farmers are simple and unaware of technologies in this field. Various methods have been developed to diagnose disease. These methods are unavailable for many farmers and need domain knowledge or a great amount and resources to carry out. This study bridges a gap between farmers and technology by using CNN in detecting diseases in plants.