Expert System for Monitoring the Malaxing State of the Olive Paste Based on Computer Vision †

Abstract

The malaxing of olive paste is one of the most important sub-processes in the virgin olive oil production process. The master continuously supervises the olive paste inside the themomixer to assess the preparation state of the olive paste and he acts manually over the process variables. The viscosity, granularity, and the presence of olive oil over the paste are the main indicators of the olive paste state. Furthermore, the temperature, time, coadjuvant addition and the shovel speeds are the process variables in the thermomixer. In this work, different image-processing parameters have been proposed to automatically assess the aforementioned indicators and they have been used as inputs in the designed fuzzy controller. Also, the outputs of this controller have been evaluated according to a sequence of images obtained inside the thermomixer and during the malaxing process in a real olive mill.

Gray-level Co-Occurrence Matrix application to Images Processing of crushed Olives fruits.

“…Uno de los métodos más comunes en el análisis de textura de imágenes es la matriz de co-ocurrencias de los niveles de grises (GLCM). Éste método, empleado por primera vez por Haralick et al [1] para caracterizar y discriminar diferentes tipos de terrenos a partir de imágenes satelitales, es muy utilizado en sectores tan diversos como la teledetección [2], análisis de superficies [3] [4], medicina [5][6] [7], últimamente en la industria agroalimentaria [8] [9][10] [11] [12] y en particular en la industria de elaboración del aceite de oliva virgen [13] [14].…”