Abstract: Rice root traits have relations with yield, which will become one of the trends for future research to improve rice varieties by root system breeding. In order to study  the relevant traits of rice root system, this paper designed a method to dynamically and non-destructively measure surface root traits of rice plants. Planting rice varieties in transparent PMMA tubes, 27 surface root traits could be obtained through combining visible light image with processing technology. The results showed that surface roots growth rate at 32~42 d reached the maximum, then they grew slow down or became dead. The centroid got down faster in 22~54 d, and then followed by little changes. The ratio between surface root area and above ground area had similar normal distribution relation with its yield. When the ratio was between 0.45~0.65, the yield was higher. The surface root coverage decreased from top to bottom, but different rice varieties had different surface roots distributions. The coverage rate of the surface root at top areas were varied from the maximum 45% to the minimum 25%. This paper preliminarily explored nondestructive measurement of rice root traits, understood in-depth rice root structure, and played certain role in promoting development of plant phenomics.

Key words: surface root traits of rice plant, nondestructive measurement, movable imaging darkroom, image processing, plant phenomics

摘要： 水稻根系性状和产量之间存在着联系，通过根系育种来改良水稻品种成为未来研究的趋势之一。为研究水稻根系相关性状，设计了一种可以动态无损测量水稻表层根系的方法。将水稻植株种植于透明PMMA管中，利用可见光成像并结合图像处理技术，计算获取27个水稻表层根系性状。结果表明，水稻表层根系32~42 d生长速率达到最大，随后生长减缓或者死亡；质心位置在22~54 d下移较快，随后变化较小；表层根系与地上部分面积之比同产量之间存在类似于正态分布的关系，比例在0.45至0.65之间时产量较高；表层根系从上至下覆盖率递减，但不同品种水稻表层根系分布不同，最上方区域的表层根系覆盖率从最大45%到最小25%。该结果对水稻根系性状无损测量进行了初步探索，并对深入了解水稻根系结构和功能，以及推动植物表型组学的发展具有一定作用。

关键词: 水稻表层根系性状, 无损测量, 移动成像暗室, 图像处理, 植物表型组学