DEVELOPMENT OF A COMPUTATIONAL TOOL TO IDENTIFY NUTRITIONAL DEFICIENCIES IN CURUBA PLANTS (Passiflora tripartita var. molissima)

Abstract

Nowadays, the determination of nutritional deficiencies in plants has become one of the most important analysis of crop growth. Often, this analysis is performed manually by specialized personnel for long period of time, leading to an error in the procedure. The purpose of this research is to develop a computational tool for the identification of nutritional deficiencies such as Nitrogen, Potassium and Phosphorus in Curuba plants. The methodology includes 3D designs, tool construction and the detection of nutritional deficiencies by image processing techniques in MatLab. Finally, the results obtained by the developed computational tool with respect to the data provided in a traditional way by the technical experts, presented an accuracy of 100%, 80%, 91% and 94% for the identification of nutritional deficiencies of Nitrogen, potassium, other type of deficiency and Phosphorus respectively. On the other hand, it provides a technological support for growers during the life cycle of the plant, determining quickly and accurately the characteristic patterns that allow differentiating a plant with nutritional deficiencies, allowing specialists to reduce time periods and perform a more objective and realiable analysis for the identification of nutritional deficiencies.