论文部分内容阅读
在林业系统注册树木种类时,通常采用大量树叶上的可见外形特点来辨别数目繁多的树木种类。针对此问题,本文提出了基于神经网络的树叶识别方法,该方法建立并管理一个分等级的树叶图像体系,通过对不同种类树叶的边缘提取来识别每种树叶的特征点,从而得到树叶的外部轮廓来区分树木种类:给出了相应的改进型神经网络算法,并给出详细的论述过程:最后,通过Java语言给出了系统实现并做了大量的数值仿真。结果证明本文方法是可行有效的。图4表1参8。
When registering tree species in the forestry system, a large number of tree species are often identified by the visible appearance of a large number of leaves. In order to solve this problem, this paper proposes a leaf recognition method based on neural network. This method establishes and manages a hierarchical leaf image system. By extracting the edges of different kinds of leaves to identify the characteristic points of each leaf, the leaf exterior Outline to distinguish the types of trees: given the corresponding improved neural network algorithm, and gives a detailed discussion of the process: Finally, the Java system is given and made a lot of numerical simulation. The results show that this method is feasible and effective. Figure 4 Table 1 参 8.