Non-destructive Mathematical Modeling Techniques for Fruit Volume Estimation: A Systematic Review and Meta-analysis

In the recent past, the global fruit industry has experienced incredible growth, fueled by growing per capita earnings and greater health consciousness for fresh produce. Fruit volume plays a key role in precise yield estimation, improving productivity, sorting and packaging. This systematic review accompanied by meta-analysis sheds light on the non-destructive techniques and algorithms used in the estimation of fruit volume through mathematical modeling. A total of 50 studies published between 2008 and 2023 were reviewed in this work in 2023 at Shivaji College (University of Delhi), Delhi. Reviewing the studies analytically, the modeling techniques adopted by researchers usually belonged to categories of either statistical modeling or geometric modeling. An I-square statistic of 88.48% was obtained in the heterogeneity analysis demonstrating the extreme diversity between the above categories. Egger’s and Begg’s tests were also performed for examining the presence of publication bias, however they did not turn up any compelling evidence of its occurrence. The comparison between different categories with their coefficient of determination (R²) between estimated and actual volume was also established using effect measures like odds ratios, risk ratiosand weighted odds ratios while sensitivity analysis was performed to assess the changes in result. This study also elucidates the strengths and shortcomings of different non-destructive techniques while using statistical methods to identify the performance of individual studies and to find the most suitable approach for estimating fruit volume. The meta-analysis concluded that the studies following statistical approach offered better R² values as compared to other methodologies.