【摘要】 玉米籽粒角质率是评价玉米品质的关键指标。针对现有检测方法误差大、通用性差或需要破坏种子等问题，以324份玉米自交系为材料，结合近红外光谱技术与线性偏最小二乘回归法，建立了玉米角质率和粉质率的无损检测模型，2种模型均具有高的预测精度和好的稳定性(R_val²>0.900,RPD_p>3.000)。研究结果为玉米籽粒角质率的快速无损高通量检测提供了可靠方法，对玉米育种和品质改良具有重要的理论意义和应用价值。更多还原

【Abstract】 The vitreous endosperm ratio of maize kernels was a key indicator for evaluating the quality of maize（Zea mays L.）. In view of the problems such as large errors, poor universality or the need to damage seeds in the existing detection methods, 324 maize inbred lines were used as materials. Combined with near-infrared spectroscopy technology and linear partial least squares regression method, non-destructive detection models for vitreous endosperm ratio and floury endosperm ratio of maize kernels were established. Both models had high prediction accuracy and good stability(R_val²>0.900, RPD_p>3.000). The research results provided a reliable method for the rapid, non-destructive and high-throughput detection for vitreous endosperm ratio of maize kernels, which had important theoretical significance and application value for maize breeding and quality improvement.更多还原