Paddy crops are one of the major sources of income in the country as rice is the main food of the population. Rice is a basic necessity and it's very high demand among consumers in the country causes farmers to use fertilizer to produce enough rice. The use of fertilizers will bring benefits to farmers but if the quantities of fertilizer are not controlled then it can cause losses. Typically, fertilizers contain nitrogen which is the source of environmental pollution if nitrogen levels are too high. The rate of use of fertilizer or nitrogen plays an important role in the process of growing paddy trees. The concentration of nitrogen levels can be determined through the colour of the leaf of the paddy tree such as yellowish green, light green and dark green. The colour indicates the concentration of nitrogen contained in the rice plant. The yellowish green colour indicates that the rice trees do not get enough fertilizer and the growth of stunted rice trees. While the dark green colour indicates high nitrogen levels, the mature process of rice trees becomes slow. The light green leafy paddy is a healthy tree and the nitrogen content contained in paddy is perfect. Traditionally, farmers identify nitrogen concentrations manually by surveying the rough eyes. However, this method requires a lot of workforces, time and not consistency. In addition, that method is impossible to get accurate results and there is an error in the survey process because different people will have a different opinion specific task. The purpose of this project is to provide an image processing system to estimate the concentration of nitrogen in paddy trees through leaf colour. Concentration is estimated by using image processing techniques, automated detection and extraction system besides data collection aims to help farmers in the future.
