Tomato Leaf Disease Detection and Classification using Convolution Neural Network
Domain:
agricultureRecord type:
paper
Creator:
Aba
Publisher:
Nat
Host:
Tomato (S. lycopersicum) is one of the most widely consumed vegetable crops in the world. It is a nutritionally well-balanced food that contains a substantial amount of vitamin A and vitamin
C, thus it plays an important role towards ensuring food security and nutrition. Globally, among horticultural products tomato ranks third for volumes of production – after potato and sweet potato and fist in terms of processing volumes. By virtue of its high yields per hectare and its processing potential, it is one of the strategic commodities prioritized by the farmer for their consumption and agro-industry. Tomato is a popular and widely grown vegetable crop in Ethiopia, ranking 8th in terms of annual national production. It is consumed in every household in different modes, but in certain areas, such as Walo, Hararge, Shawa, Jimma and Wallaga, it is also an important co-staple food [1]. Tomatoes grows on mostly any well drained soil and Nine out of 10 farmers grow tomatoes in their field. Many gardeners also grow tomatoes in their garden to use fresh grown tomato in their kitchens and get a good taste of food. However, farmers and gardeners are sometimes unable to get proper progress of the plant growth. The tomatoes may not sometime appear on plant or sometimes the tomatoes may get bad looking and disease-full black spots at bottom part. The identification of tomato plant disease may start from, to diagnose the portion of pant leaf having infection then to note the differences such as brown or black patches and holes on the plant and then to look for the insects also. The tomato problems may be divided into two sections: bacteria or fungi or poor cultivation habits causing 16 diseases while insects causing 5 other types of diseases [2]. Plant disease recognition using images from digital and mobile phone camera proves to be a significant challenge. The recent trend in use of various machine learning algorithms for plant disease classification has shown promising results in few selected diseases and crops[3]. Evolution of deep Convolutional Neural Network (CNN) based architectures have further enhanced the accuracy of classification significantly[4].
C, thus it plays an important role towards ensuring food security and nutrition. Globally, among horticultural products tomato ranks third for volumes of production – after potato and sweet potato and fist in terms of processing volumes. By virtue of its high yields per hectare and its processing potential, it is one of the strategic commodities prioritized by the farmer for their consumption and agro-industry. Tomato is a popular and widely grown vegetable crop in Ethiopia, ranking 8th in terms of annual national production. It is consumed in every household in different modes, but in certain areas, such as Walo, Hararge, Shawa, Jimma and Wallaga, it is also an important co-staple food [1]. Tomatoes grows on mostly any well drained soil and Nine out of 10 farmers grow tomatoes in their field. Many gardeners also grow tomatoes in their garden to use fresh grown tomato in their kitchens and get a good taste of food. However, farmers and gardeners are sometimes unable to get proper progress of the plant growth. The tomatoes may not sometime appear on plant or sometimes the tomatoes may get bad looking and disease-full black spots at bottom part. The identification of tomato plant disease may start from, to diagnose the portion of pant leaf having infection then to note the differences such as brown or black patches and holes on the plant and then to look for the insects also. The tomato problems may be divided into two sections: bacteria or fungi or poor cultivation habits causing 16 diseases while insects causing 5 other types of diseases [2]. Plant disease recognition using images from digital and mobile phone camera proves to be a significant challenge. The recent trend in use of various machine learning algorithms for plant disease classification has shown promising results in few selected diseases and crops[3]. Evolution of deep Convolutional Neural Network (CNN) based architectures have further enhanced the accuracy of classification significantly[4].