摘要
提取植物叶片的颜色特征是LED植物生长柜实现智能监控的关键之一。针对植物生长柜,提出了一种用于检测叶片是否发生病变的简便方法。通过计算机提取叶片在RGB和HSV颜色系统下的颜色特征,对比选择合适的颜色参数以及计算颜色参数的均值和方差,分别从整体和部分分析判断叶片是否健康、是否发生病变。将采集的健康叶片图像及病变叶片图像进行对比测试,实验结果表明:色调H参数能够实现较好地反映出叶片的颜色特征和生长的健康状态,而且能够将病变部分区分并提取出来,以此可以诊断病因,为植物生长柜的管理者尽早发现问题来采取合理的措施提供理论依据。
The color feature extraction of plant leaf is one of the key factors to realize the intelligent control of the LED plant growth cabinet. For the plant growth cabinet, we propose a simple method to identify whether the leaves are lesion. The leaf color feature was extracted in the RGB and HSV color system by the computer, the appropriate color parameters were chose to calculate the mean and variance, and separately from the whole and part, the leaf was analyzed to determine whether it is healthy or diseased. Compare the health and ill leaf image, the H parameter can better reflect the color feature and the health status of the blade and it can extract the lesion part. It is easy to diagnose the cause which provides theoretical basis for taking reasonable steps for the management of the plant growth cabinet.
出处
《中国农机化学报》
2015年第1期107-110,共4页
Journal of Chinese Agricultural Mechanization
基金
国家自然科学基金项目(61178048)--高速率并行光互连器件的基础科学问题研究
国家社会科学基金项目(BFA110049)--以机器人教育为载体的高中与大学创新教育有效衔接的实践研究
关键词
植物生长柜
叶片
颜色特征
H参数
病变检测
plant growth cabinet
leaf
the color feature
H parameters
lesion detection
