| 摘 要: 针对自然环境下的叶片图像分割，提出了一种基于支持向量机的叶片图像分割算法。该方法首先将图像少量像素点分别标记为叶片前景样本和叶片背景样本，然后根据样本数据建立支持向量机分类决策模型，最后根据预测模型对整个图像像素点进行分类，将叶片图像从背景中分割出来。实验结果表明，该方法能够对含有反光区域的叶片实现准确分割，相比基于聚类的叶片分割算法分割精度更好，算法耗费时间更短。 |
| 关键词: 叶片分割；支持向量机；自然环境；反光区域 |
|
中图分类号: TP391.41
文献标识码: A
|
| 基金项目: 安徽省高校自然科学重点研究项目(KJ2020A0392)；安徽中医药大学校级自然重点项目(2020zrzd16)；安徽中医药大学校级自然一般项目(2020zryb09). |
|
| Leaf Image Segmentation based on Support Vector Machine |
|
WU Zhangqian, WANG Qing
|
(College of Medicine Information Engineering, Anhui University of Chinese Medicine, Hefei 230012, China )
wzq6529@ahtcm.edu.cn; wangqing@ahtcm.edu.cn
|
| Abstract: For leaf image segmentation in natural environment, this paper proposes a leaf image segmentation algorithm based on support vector machine. Firstly, a small number of image pixels are marked as leaf foreground samples and leaf background samples. Then, a classification decision model of support vector machine is established according to the sample data. Finally, the entire image pixels are classified according to the prediction model, and the leaf image is segmented from the background. Experimental results show that the proposed method can accurately segment the leaves with reflective areas. Compared with the clustering-based leaf segmentation algorithm, it has better accuracy in segmentation and takes less time. |
| Keywords: leaf segmentation; support vector machine; natural environment; reflective area |
