In this work, a correlation model of tilapia fillets quality evaluation was established by principal component analysis (PCA) based on fractal dimension change. The tilapia fillets were stored at 3 and −2°C, respectively, and the pH, color, electrical conductivity (EC), total volatile base nitrogen (TVB‐N), texture, and fractal dimension were periodically measured. A major component was extracted and it explained 82.726% of the total variance. Then the comprehensive quality score was calculated during the storage of the fillets. Three function correlation models including linear function, logarithmic function, and quadratic function were obtained by establishing the relationship between the comprehensive quality score of fillets and the fractal dimension. The determination coefficient R² of quadratic function correlation model reached 0.841 and 0.962 under 3 and −2°C, respectively. This result indicated that the quality evaluation was accurate under low‐temperature storage conditions by quadratic function correlation model based on the change of fractal dimension.

中文翻译：

---

基于分形维数变化在3和-2°C下储存的罗非鱼片的质量评估方法

本文通过分形维数变化的主成分分析法建立了罗非鱼片品质评价的相关模型。罗非鱼片分别保存在3和-2°C下，并定期测量pH，颜色，电导率（EC），总挥发性基础氮（TVB-N），质构和分形维数。提取了一个主要成分，它解释了总方差的82.726％。然后在圆角存储期间计算综合质量得分。通过建立圆角的综合质量得分与分形维数之间的关系，获得了三个函数相关模型，包括线性函数，对数函数和二次函数。判定系数R ²在3°C和-2°C下，二次函数相关模型的均值分别达到0.841和0.962。该结果表明，基于分形维数变化的二次函数相关模型，在低温存储条件下质量评估是准确的。