中文摘要：

为了掌握智能植物生长柜中小白菜的成熟情况,便于对柜内环境参数实现智能控制,提出了利用小白菜的外部形态特征,特别是提取根系形态特征并将其与地上部分形态特征相结合来检测小白菜成熟度的方法。通过Matlab图像处理工具箱对采集的小白菜图像进行阈值分割和特征提取,然后将小白菜上、下两部分的形态特征数据作为训练样本,分别建立对应的神经网络成熟度检测模型,并将神经网络检测值利用贝叶斯理论来对其进行信息融合,从而进一步提高神经网络模型检测的准确性。

英文摘要：

In order to better grasp the maturity of pakchoi in the plant growth cabinet to realize the intelligent control of the environment in the cabinet, the method of detecting the pakchoi maturity was proposed by using the external morphological characteristics of pakchoi ,especially the extraction of root morphological characteristics, and combining the above ground morphological features with the root morphological characteristics. The threshold segmentation and feature extraction were conducted for the gathered pictures of pakchoi by the image processing toolbox of Matlab, and then the data of the above ground morphological features and the root morphological characteristics of pakchoi were taken as training samples to establish two neural network maturity detection modle of the ground and underground. Finally, the information fusion of the maturity by Bayesian theory was done, and which could further improve the detection accuracy of the neural network model.