利用G/R分量比值图分割水稻植株图像

Detection of paddy based on ratio image of G component to R component

在非实验室的环境下,利用水稻作物植株的RGB彩色图像计算的R分量与G分量的比值图分割水稻植株。比值图中水稻植株与周围环境的差异比较大,利用这一特点,用阈值分割的方法对图像进行分割,提取水稻植株,结果比较理想。对于大量图片水稻植株的提取,使用迭代的方法自动选择最佳的分割阈值,分割成功率较高。

In non-lab environment, ratio image of G component to R component of RGB image of paddy is used to detect the paddy body. In ratio image, the paddy is much different to the environment. With this characteristic, thresholding is applied to the segmentation of paddy image. Iteration is used to select the best thresholds in the segmentations of a great deal of paddy images and the ratio of successful segmentation is high.