Aquaculture requires precise non-invasive methods for biomass estimation. This research validates a novel computer vision methodology that uses a signature function-based feature extraction algorithm combining statistical morphological analysis of the size and shape of fish and machine learning to improve the accuracy of biomass estimation in fishponds and is specifically applied to tilapia (Oreochromis niloticus). These features that are automatically extracted from images are put to the test against previously manually extracted features by comparing the results when applied to three common machine learning methods under two different lighting conditions. The dataset for this analysis encompasses 129 tilapia samples. The results give promising outcomes since the multilayer perceptron model shows robust performance, consistently demonstrating superior accuracy across different features and lighting conditions. The interpretable nature of the model, rooted in the statistical features of the signature function, could provide insights into the morphological and allometric changes at different developmental stages. A comparative analysis against existing literature underscores the competitiveness of the proposed methodology, pointing to advancements in precision, interpretability, and species versatility. This research contributes significantly to the field, accelerating the quest for non-invasive fish biometrics that can be generalized across various aquaculture species in different stages of development. In combination with detection, tracking, and posture recognition, deep learning methodologies such as the one provided in the latest studies could generate a powerful method for real-time fish morphology development, biomass estimation, and welfare monitoring, which are crucial for the effective management of fish farms.

Pond biomass estimation and non-invasive biometrics are necessary problems to solve in shrimp farming to achieve the optimization of currently outdated manual processes which are slow, inaccurate, imprecise, and prone to errors. This dataset was collected to develop and test computer vision and artificial intelligence models to accurately detect shrimps and estimate their biomass. The dataset was collected in three ponds, two in an industrial farm and the other in a university pond cultivated for academic purposes. 170 shrimps were sampled by taking pictures and manual measurements of their total length, cephalothorax length, and weight. A total of 5507 shrimp’ images were taken which were put in containers equipped with cameras and with a water level of 10 cm. The dataset is organized into five sub-datasets folders and excel files containing the manual measurements taken with a scale and vernier from each sample. This dataset could be used to compare and develop different detection and biomass estimation computer vision models since it presents a good amount of images and samples of shrimps cultivated in different conditions which can allow models to relate image features of shrimp samples with their corresponding weight and also compare these models against the machine learning models that can be applied solely to the manually extracted features stored in the excel files for biomass estimation.