Non-destructive measurements for the evaluation of beef freshness based on airflow interaction and a structured light 3D imaging technique and comparison with pH value and total volatile basic nitrogen content

This study aimed to investigate the potential ability to predict total volatile basic nitrogen (TVB-N) content and pH value for the freshness evaluation of beef by examining instan-taneous elastoplastic deformation of the meat caused by airflow using a structured light 3D imaging technique. A constant airflow impacted the beef surface whilst a 3D camera continuously captured point cloud images of the concave area. The volume, surface area, and depth of spatial deformation, as well as the projected area of three planes, were ob-tained using the K-nearest neighbour algorithm, down-sampling processing, the oriented bounding box algorithm, and the Alpha Shape algorithm. The instantaneous recovery and residual of each quantified deformation feature were obtained at the moment of air unloading to get the instantaneous elastic-plastic deformation and degree of instanta-neous elastic-plastic deformation of beef's spatial deformation feature. A correlation analysis of the obtained features and two freshness indicators (pH value and TVB-N con-tent) was carried out for feature selection and modelling. The results showed that the models using the degree of instantaneous elastic deformation of spatial deformation fea-tures of beef with feature selection produced the best prediction results in this study. For beef pH values and TVB-N content evaluation, the correlation coefficient in the prediction set were 0.745 and 0.844, respectively.(c) 2023 IAgrE. Published by Elsevier Ltd. All rights reserved.

Automated summary

This study used a structured light 3D imaging technique to investigate the potential to predict total volatile basic nitrogen (TVB-N) content and pH value for evaluating the freshness of beef. By examining the instantaneous elastoplastic deformation of the meat caused by airflow, the study obtained various deformation features of the beef's spatial deformation. Correlation analysis and modeling were conducted to select the most effective features and develop prediction models. The results showed that the models using the degree of instantaneous elastic deformation of spatial deformation features produced the best prediction results, with correlation coefficients of 0.745 and 0.844 for pH value and TVB-N content evaluation, respectively. (c) 2023 IAgrE. Published by Elsevier Ltd. All rights reserved.