图像分割技术之迭代算法在杂草识别中的应用

The Application of Iterative Algorithm of Picture Division in Weed Recognition

计算机图像处理是指将图像信号转换成数字信号并利用计算机对其进行处理的过程。图像分割技术是图像处理之基本技术,图像分割就是在一幅影像中查找同类的区域。图像分割的应用范围涵盖了图像噪声的过滤以及特征的提取和识别。在介绍图像分割的主要特征的基础上,针对农田杂草识别中图像分割特点,提出了应用迭代算法的分割方法,并开发了相应的软件。实验验证该方法用于农田杂草识别前的图像的二值化处理,能够很好地消除图像噪声,得到连续的图像边界,可以有效地实现图像的二值化处理,为后续的杂草识别打下良好的基础。

The computer picture processing is the process that the picture signal transforms to the digital signal and it is handled by the computer. The picture division is one kind of basic picture processing technology and is the process that the similar regions are searched in an image. It includes the picture noise filtering as well as retrieval and recognition of characteristics. On the basis of introduction of the main characteristics of the picture division and those in farmland weed recognition, iterative algorithm was put forward and the corresponding software was exploited. The experiment confirms the picture noise can be eliminated effectively and continuous image border can be received if the method is used in the two-valued image prosessing before recognition. The results lay a sound foundation for the further weed recognition research.