基于Kinect传感器的苹果树冠层形态信息测量方法研究

Research on Measurement Method of Apple Canopy Morphological Information Based on Kinect Sensor

在现代化果业中,建立一套完整、精准、自动化的果树冠层形态信息感知系统,可以为现代标准化果园智能化管控提供科学的数据支持。为了能够便捷、高效地获取苹果树冠层形态结构以及果实生长信息,设计了一种基于Kinect传感器的苹果树冠层形态信息测量方法,实现对苹果树冠层原始图像信息的感知、三维点云的建立、以及形态参数的提取。研究结果表明:采用水平视角度测量苹果树冠层宽幅、高度信息相对误差值最小,分别为2.91%和3.68%。本研究提出了苹果树冠层形态信息测量方法,并分析了平视、仰视、俯视三个角度的测量准确性,对促进我国现代标准化果园信息化与智能化管理具有重要的意义。

In the modern fruit industry, the establishment of a complete, accurate and automatic fruit tree canopy shape information perception system can provide scientific data support for modern standardized orchard intelligent management and control. In order to obtain the information of apple tree canopy morphology and fruit growth conveniently and efficiently, a method of apple tree canopy morphology information measurement based on Kinect sensor was designed. The perception of the original image information of apple tree canopy, the establishment of three-dimensional point clouds and the extraction of morphological parameters are realized. The results showed that the relative errors of apple tree canopy width and height measured fromahorizontal angle were 2.91% and 3.68%, respectively. In this study, the measurement method of apple tree canopy morphology information was put forward, and the measurement accuracy of heads-up, up-looking and overlooking was analyzed, which is of great significance to promote the information and intelligent management of modern standardized orchards in China.