Physicochemical quality visual characterization postharvest of bell pepper applying recent mathematics techniques in the digital image processing

Abstract

Mexico is one of the main countries in the production and exportation of bell pepper (Capsicum annuum) hence, emerging techniques to monitor and assess quality parameters postharvest is a constant topic of interest. On the other hand, color image processing is a non-invasive technique that is gaining popularity in the agro-industrial sector. Color analysis helps as an indirect or correlation measure to quantify and qualify visual quality attributes (size, shape, maturity, etc.) and even non-visual ones such as flavor. Recent techniques involve a pre-processing of the information (changing the RGB space into another) in order to obtain the chromaticity and luminescence parameters that point to the characterization of the physicochemical profile; this stage may present a loss of information. Therefore, this work presents the use of Quaternion Algebra for the management of information as a single entity, avoiding the pre-processing stage. The image is transformed into a pure quaternion; processed using the Fast Fourier Transform described in Quaternions. Image acquisition is carried out using any mobile device or low-cost sensor, which provides an advantage over other proposals. The results show that the determination of physical measurements such as length, width and thickness are characterized according to pre-established standards. In addition, we identify the state of maturity; as well as areas in which damage or injuries are found. The validation of the results obtained was done in consideration of Good Management Practices as well as, the state of the art having a high agreement with them.