Measurement of Morphometric Traits and Estimation of Body Weight of Gilthead Seabream and European Seabass using 3D Reconstruction Simulations with the Coded Structured Light Technique

Estimating fish biometric characteristics without human intervention is one of the most significant challenges in modern aquaculture. In this paper, a simulation-based method for estimating key geometric characteristics and the weight of gilthead seabream and European seabass is presented. The proposed method implements 3D reconstruction of these fish using the Coded Structured Light method, which is an efficient stereoscopic technique. The extracted 3D reconstructions are used to calculate the standard length, body height and area of the fish, and subsequently, based on these measurements, to estimate their weight. The experiments are conducted by placing the fish at various distances from the system components during the simulations. Obtained results indicate that depending on measuring distance and fish weight class, precise measurements of morphometric traits, as well as accurate body weight estimation through regression analysis, is achieved (\({R}^{2}>0.99\), MAPE as low as 4.9%). Overall, the proposed methodology demonstrates remarkable effectiveness, encouraging enough to suggest that the presented technique can serve as the initial stage for similarly efficient evaluations in the field.