Biometrical Optimization System for Red Tilapia (Oreochromis sp.) through Image Processing in Growing Environments

Abstract

The costs of an aquaculture production process are affected by about 70% by the technical food management, in that sense, it is absolutely necessary to accurately establish the food rations throughout the cultivation cycle, making adjustments weekly, so the determination of biometric variables is required with the same periodicity. The objective of this research was to make possible an alternative for obtaining biometric information in an aquaculture production process, obviating harmful aspects, which occur in the development of traditional biometrics, where it is required to perform invasive procedures (partial fishing) that generate negative impacts on cultured organisms. In this sense, in the development of the present study, magnitudes of the specimens of red tilapia (Oreochromis sp) were determined, through the analysis of images obtained in culture environments at high densities, all this, using a software specially designed to measure the length of the specimens, managing to generate a lower impact on the fish. For the development of the software it became necessary to determine the events and sequences, thanks to unified modeling language diagrams - UML, in this specific case the use cases, sequence and entity relationship were used. Subsequently, each of the events and sequences determined in the UML were embodied in code, using the C# programming language and the Visual Studio (IDE) in its community version. On the other hand, the calibration of the system was carried out in three clearly defined phases: 1. selecting the image capture device, 2. support structure adjustment of the image capture equipment, and 3. calibration of the measurement software.