基于时序巡航图像的茶树生长监测研究

茶树在生长过程会发生各种变化,对茶树进行实时监测有助于及时发现其生长异常状况,并采取有效手段减少经济损失、提高茶叶品质和产量。该研究设计了1个茶树生长图像监测系统,在茶园中架设1台可变焦云台摄像机,将茶园划分为若干监测区域,利用摄像机的变焦和转动获取监测区域的视频流,实现茶树生长的实时监测。定时对各监测区域进行图像抓拍,并按监测区域和时间保存到茶树生长图像库。通过对茶树生长图像库的分析,可以得到茶树的嫩芽萌发情况、病害发生过程、营养缺失情况等生长信息,以及光照、天气变化等环境信息。通过茶树图像的颜色特征参数与含氮量之间的回归分析,得到茶树含氮量的快速监测模型,平均相对误差在6%以内。本系统为获取茶树生长信息提供了一种快捷、简便、经济的技术手段,同时也适用于其他作物的生长监测。

Tea plant growth monitoring based on time series cruise images

The growth of tea plant is easily affected by many factors, such as soil conditions, diseases and insect pests, pruning and picking, it is necessary to monitor the growth of tea plant. However, manpower monitoring is time-consuming, laborious and inefficient. In order to solve this problem, a tea plant growth image monitoring system was designed in this paper, which could monitor tea plant in real time. It was helpful to find out the abnormal situation of tea plant growth in time, and take effective measures to reduce economic losses and improve the quality and yield of tea. In the proposed design, a zoom camera was set up at suitable position, and the tea plantation was divided into several monitoring regions. The video flow of the monitoring region was obtained by zoom and rotation of the camera to accomplish real-time tea plant monitoring. Images of each monitoring region were captured in sequence and saved into image database of tea plant growth according to the monitoring region and time. By analyzing the image database, growth information such as the germination of tea buds, occurrence process of diseases, nutrition, and environmental information such as light, weather change, was obtained to understand the growth of tea plant in a more comprehensive way. Nitrogen diagnose method was also proposed to get real-time nutrition of tea plant. Through regression analysis between color parameters and nitrogen content, a fast monitoring model of nitrogen content of tea plant was obtained, with mean relative error less than 6%. The design provided a fast, simple and economical mean to obtain growth information from tea plant and other crops.