Low-temperature stress is one of the factors affecting the growth and development of peanuts. Currently, biochemical detection technologies for crop freeze damage are well established. In the context of rapid development in optical sensing technology and smart agriculture, non-destructive crop freeze damage detection based on such technologies has gained increasing attention. The accurate detection, early warning, and targeted control of crop cold damage are particularly important. In this study, 70 peanut germplasm resources were collected and used for the research objectives. Indoor low-temperature seedling identification was conducted at 25 °C (the control group) and 5 °C (low-temperature stress group) for 7 days. Photosynthetic fluorescence values in leaves, as well as 13 indicators (Fo, Fm, Fv, Fv/Fm, Fv’/Fm’, ΦPSII, NPQ, qP, Rfd, Pn, Gs, Ci, and Tr), were analyzed for their responses to low-temperature stress. The results showed that under low-temperature stress, the Pn and Ci of peanut seedlings exhibited an ascending trend, while Tr and other indicators showed a decreasing trend compared to the control group. Based on the relative coefficients of resistance to low temperature for each individual indicator, a comprehensive non-destructive evaluation of cold resistance was conducted using methods such as principal component analysis, cluster analysis, and stepwise regression. Through principal component analysis, the 13 individual physiological indicators were transformed into 3 comprehensive indicators. The 70 peanut varieties were divided into 4 categories based on their resistance to low temperature: sensitive materials, moderately sensitive materials, moderately cold-tolerant materials, and cold-tolerant materials. Additionally, a mathematical model for evaluating cold resistance in peanuts was established.