ANÁLISE DE TÉCNICAS DE LIMIARIZAÇÃO EM PROGRAMAS DE ANÁLISE DE CARTÕES HIDROSSENSÍVEIS

Abstract

The effect of environmental contamination by pesticides is becoming a major concern and a frequent topic of debate worldwide. Optimize the process of applying the pesticide, is the most efficient approach found so far to reduce environmental pollution without sacrificing effectiveness in controlling plant diseases. Among the various techniques available for quality control in the process of application of pesticides, the use of water sensitive cards is one of the most popular, but the analysis of these cards is very difficult. There are some computer programs for analyzing images of water sensitive cards, but none is available for access online. Thus, this study aimed to test thresholding techniques for developing an online application for analyzing water sensitive cards. We tested techniques for global thresholding (Otsu´s method), adaptive local thresholding (Wellner’s method) and multithresholding (Bernsen and Papamarkos’ methods). The Otsu method, which uses the technique of global thresholding produced the best results and was chosen to be used in our application online.