基于计算机视觉的植物水分胁迫状况监测方法

水分胁迫是影响植物生长最为普遍的环境威胁。水分亏缺的监测对设施农业、精确农业的发展具有重要意义。植物本身具有一种适应土壤水分胁迫的生理调节机制,其中叶片是植物外部形态中反应最为敏感的器官,这就为水分亏缺诊断提供了信号和依据。根据植物叶片对水分胁迫反应敏感这一特点,提出一种基于计算机视觉的植物水分胁迫状况无损监测新方法。首先根据环境情况对植物图像进行预处理;然后针对光照不均匀、阴影等造成的分割困难,采用了一种局部阈值图像分割算法,得到植物叶片的图像;最后提取植物叶片的叶尖相对距离和叶尖倾角这两个形态参数,实现植物水分亏缺程度的连续无损监测。实验结果表明,该方法监测的叶尖相对距离和叶尖倾角的变化趋势能够较好地反映出植物的水分亏缺程度,对于水分胁迫状况的监测,具有较高的应用价值。

Monitor method of plant water stress based on computer vision

Water stress is the most popular environment menace that affects plant growth. Monitor of water deficit is very important to the facility agriculture and precision agriculture. In plants, there is a physiological mechanism of adaptability to soil water stress, and leaf is the most sensitive organ of the plant external morphology, which provides the signal and evidence of water stress. On the basis of the characteristic that the plant leaf is sensitive to water stress, the paper presents a new non-contact method to monitor the water stress of plant by computer version. First, preprocess the image according to the environment condition; then use a local threshold image segmentation algorithm to solve the segmentation difficulty caused by uneven illumination and shadow; at last, extract two morphological parameters of plant leaf: tip relative distance and tip angle, and achieve the continuous non-contact monitoring for water stress level of plant. The experimental results show that the trends of tip relative distance and tip angle monitored by this method can clearly reflect the water deficit level of plants, which has a high application value in monitoring the water stress.