Mobile application for determining input level of fertilizer and detecting diseases in crops
MetadataShow full item record
Economic development and food security in Bangladesh significantly depend on agriculture production. Despite progresses made in agricultural techniques, farmers in Bangladesh still face difficulties in determining required levels of fertilizer and detecting different diseases during cultivation period. Using adequate fertilizer during specific periods of production and early detection of diseases ensure effective and efficient use of fertilizer and good harvest of crops. On the other hand, mobile phone usage has increased exponentially among the population of Bangladesh.. People from all walks of life are using mobile phones and different associated applications for gaining economic and social benefits. However, very few mobile phone applications benefit agricultural production and specifically aim farmers. This thesis is aimed at developing an android application to generate an automated mobile-based system to serve two purposes: a) determine the amount of input level of fertilizer for paddy, and b) detect diseases of potato through image processing of leaves. The purpose of such development was to assemble an application with a user friendly interface and implement some effective algorithms considering the problems. In order to develop the application, different algorithms, approaches and languages were examined that can ensure a user friendly interface. Determining the input level of fertilizer for paddy was executed in android using “Principal Component Analysis” algorithm. On the other hand, two specific potato leaf diseases― “Early Blight” and “Late Blight”― were detected employing “Mathematical morphology”, “Leaf Vein Detection” and “Blob detection” algorithms in MATLab. The result showed that color of paddy leaf matched precisely with the LCC to determine the input level of Urea while the application successfully recognized affected parts in the potato leaf. Evaluating with the real time data, this thesis can be a very promising application to help the farmers with their harvesting procedure.