基于图像处理技术的赤潮生物自动识别

The Automated Detection of Harmful Algae Using Image Processing

对赤潮生物进行显微自动摄像,根据不同赤潮生物显微图像的特点,研究图像分割,特征提取和分类器设计的处理过程。实验结论表明系统对赤潮生物的识别与人工识别比对,识别符合率大于80％。

Aiming at the characteristic of the different harmful algae's micro-image,this paper studied the seg- mentation, feature extraction and classification of the harmful algae's image. The result of the experiments indi- cated that the classification accuracy was up to 80%,compared with the artificial identification.