一种基于图像处理的玉米子粒饱满度测评新方法

New Method for Quantifying and Evaluating Kernel Fullness in Corn by Image Process

玉米饱满度指数定义为子粒饱满部分长度与总长度之比。以8个玉米品种各400颗典型子粒为材料,用平板扫描仪采集其非胚面的RGB图像,用ImageJ软件实现了半自动测定上述二者的长度。参试玉米品种的饱满度指数符合正态分布,均值为0.46～0.58,品种之间差异显著,变异系数为12.5%～18.8%。根据饱满度指数将8个品种聚为3个类群,与实际情况基本吻合。饱满度指数能够定量描述玉米的子粒饱满程度,可以通过图像处理软件快速准确地进行测定,为玉米品质评价和子粒生长诊断提供参考。

There is requirement for a definition of kernel fullness in cereal crops in order to provide a practical tool and a theoretic basis for kernel quality evaluation.Kernel fullness index in corn（Zea Mays L.） is defined as a ratio of full filled segment length to whole kernel length.RGB images of non-embryo sides of 400 typical kernels were scanned by a flat scanner from each of 8 corn cultivars.By means of ImageJ software,two previous length measures were obtained interactively.Kernel fullness index in corn followed the normal distribution.Cultivars tested statistically differed in kernel fullness index,of which its values ranged from 0.46-0.58.Variation coefficients varied from 12.5%-18.8%.Tested 8 cultivars were classified into 3 categories based on kernel fullness index,which coincided with field observation facts.We proposed kernel fullness index could quantify kernel fullness in corn,and proved to be easily and accurately measured using the image process software,and will provide a feasible tool for diagnosing kernel filling and commercial kernel quality.