Rapid quality assessment of salted kimchi cabbage through near-infrared spectroscopy

Due to the transition from traditionally homemade to commercially produced salted kimchi cabbage (SKC), SKC faces significant quality inconsistencies, threatening consumer satisfaction and product standardization. In this study, we introduce a novel, non-destructive method using near-infrared spectroscopy to address this issue and predict essential quality attributes of SKC and enhance quality control by enabling rapid assessment of changes in water weight, salt weight, and Young's modulus. Initial regression models were used to elucidate relationships among SKC quality attributes, while exponential models exhibited a good fit [ratio of prediction to deviation (RPD) > 2]. Subsequent prediction models, developed through various spectral pre-processing methods, highlighted the efficacy of standard normal variate pre-processing, achieving R2 values > 0.87 and RPD > 2.5 for all quality attributes. Simplified models were formulated using multiple linear regression after spectral variable selection through the successive projections algorithm, reducing spectral variables to <= 10. These models show promising applicability for enhancing SKC quality management. Future work will explore real-time in-line quality prediction through hyperspectral imaging to further streamline quality assessment processes in SKC production.