Nur Azwani Zaini, Hayatun Syamila Mamat and Suhaila Abd Halim
Keywords: semi-automated, MATLAB GUI, paddy leaf, image processing, support vector machine
Abstract: Conventionally, diseases in paddy are detected by paddy farmers manually. In manual detection, a lot of time and energy needed to detect the diseases and it is hardly possible to accurately estimate the infected areas especially in large-scale farming. This project aims to develop semi-automated system for paddy leaf disease recognition. The semi-automated system is named as i-PaddyL. It is an improved version of system consisting recognition with prevention suggestions according to leaf diseases. It also able to simplify life with the use of technology. It is believed that the system is applicable as it able to improve manual process, inexpensive, low technical management of system is required, and the results are also reliable. The system is developed using MATLAB GUI. As set of images consists of paddy leaves with three types of diseases are used for training and testing to measure the accuracy of the developed algorithm used in the system. The algorithm using image processing techniques and support vector machine as recognition process. In addition, six (6) features are calculated for each disease. Prevention suggestion is listed according to the detected disease. The suggestion can be used by farmer to think of the next step that need to take. It is found that the developed system able to calculate disease’s features and recognise the three types of diseases.
