Abstract: 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.

Key words: tea plant, video surveillance, crop growth monitoring, image acquisition