Diseño de un sistema de visión por ordenador para la caracterización del color del tomate (Solanum lycopersicum L.)

2017 
The external appearance of food is the first attribute in which consumers set their eyes on. The color, in particular, has a special relevance in the tomato (Solanum lycopersicum L.). Objective color measurement can be carried out by various instruments. These, generally, have a high-price which adds to other drawbacks for the analysis of food products. For this reason, in recent years, the use of computer vision systems (CVS) has increased. These consist of an image capturing system and a software. The latter can be an impediment for some potential users, since there are few simple computational tools for information processing and the software development is a limiting restriction for many users. In the case of the tomato, there is a specific program for color measurement, however, it cannot be modified for the incorporation of a lacking function: the analysis of the external characteristics of the fruit. Therefore, it seems opportune to develop a tool that allows analyzing the external color of the tomato in a simple, objective, efficient way and with a lower price than traditional measuring instruments. The main objective of this work is to construct a low cost CVS to measure the external color of tomato. For this purpose, an image capturing system has been designed and a program has been developed, using the software LabVIEW, for the acquisition, processing, analysis and storage of images, as well as a simple graphical environment for user manipulation. Finally, the effectiveness of the CVS has been proven by the study of a tomato fruit. The first results achieved have been the image capturing system, the program for image manipulation and the graphical user interface. The color of various colored cards has been analyzed using the CVS. These measurements have been compared with the readings made by a commercial system, the Konica Minolta CR-400 colorimeter, obtaining a standardized average error of the CVS of 2.50 %. This result demonstrates the accuracy of the built CVS.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    0
    References
    0
    Citations
    NaN
    KQI
    []