Digital Agriculture Expert Systems Using Image Processing and Machine Learning

Paddy fields have been used for cultivation over thousands of years. But

these paddy fields don't care much for replenishing. It has degraded the quality and

productivity of the plants, fruits, and also led to the depletion of the soils. This is a

serious problem that can be solved by using proper farming and technology. Most of

the farmers often used traditional farming technology such as sickle, plow, spade, weeder, weed, rake, and ax, etc. [1]. The traditional system takes more time and

human efforts to do farming [2]. But modern technologies help farmers to enhance the

quality and productivity of farming. The urban farmers avail the facilities of modern

technologies and laboratories, but the rural farmers often face difficulties due to the

non-availability of modern farming technologies, and laboratories. The productivity of

farming depends on the selection of physical and chemical properties of soil, selection

of seed, a good amount of plant photosynthesis, and early detection of plant diseases,

etc. These are the serious issues of farming.

The developed system helps the farmer to classify the best soil for Assam citrus, predict the suitable soil pH for citrus

farming, estimate the leaf chlorophyll for citrus health issues, and detect the citrus leaf

diseases at the vegetative stage.

The system shows the application of Linear Regression (LR), Artificial Neural

Network (ANN), K Nearest Neighbour (KNN), Support Vector Machine

(SVM), and Convolution Neural Network (CNN) in precision farming for the

better yields of the citrus farming.