摘 要

本文基于机器视觉开发了一套计算机/叶片交互系统，实现根据叶片几何特征值的不同而进行自动识别叶片种类，实现快速地对叶片种类区分。

本文采集了五组不同的树叶。基于HALCON图像处理软件，通过对颜色，形状以及大小进行分析。通过不同树叶的几何特征不同，选取树叶面积、圆度、充实度、周长4种几何特征值进行分析。将任意两组特征值数据合并处理并做成散点图像，观察不同树叶的集中分布位置，从而将每种树叶从其他树叶中挨个区分出来，利用HALCON软件的算子函数达到分别各种树叶的识别。

对于HALCON软件的程序进行了一次改进，对最初的程序进行了改动，使得第二次的程序变得简便。并且程序多次改进后树叶的正确识别率达到了99%，表明该方法能够成功的对树叶的识别。

关键词：树叶识别；HALCON;图像处理

Automatic identification of tree species based on leaf color image

abstract

In this paper, a set of computer - blade interaction system based on machine vision is developed to realize automatic recognition of leaf types according to the different values of the blade geometry.

In this paper, five different sets of leaves were collected. Based on HALCON image processing software, through the color, shape and size of the analysis. Based on the geometric characteristics of different leaves, the leaf area, roundness, filling degree and circumference of the 4 kinds of geometric characteristics were analyzed. The arbitrary two groups of feature value merge data processing and make scatter image, to observe the different leaves distributed locations, and each leaves from the other leaves one by one to distinguish and use of Halcon software operator function to achieve recognition of all leaves respectively.

For the HALCON software program has been an improvement, the initial program has been changed, so that the second program is simple. And the correct recognition rate of the leaf was 99% after several times, which indicated that the method was successful in the identification of leaves.

Key words: leaf recognition; HALCON; image processing

目录

1 绪论........................................................................1

1.1 目的及意义.............................................................1

1.2 国内外研究现状.........................................................2

1.3 研究内容与技术路线.....................................................3

1.4 本文结构安排...........................................................4

2 图像采集处理................................................................6

2.1 图像特征...............................................................6

2.2 图像识别与处理.........................................................6

2.3 HALCON软件.............................................................9

2.3.1 HALCON软件简介....................................................9

2.3.2 HALCON软件功能....................................................9

2.3.3 HALCON软件应用...................................................10

2.3.4 选取HALCON软件的主要理由........................................11

2.4 图像采集..............................................................11

2.5 图像处理..............................................................18

3 数据分析与程序编制.........................................................25

3.1 对树叶单个几何特征值的分析............................................25

3.2 对树叶两个个几何特征值的分析..........................................27

3.3 程序的制作............................................................33

3.4 检验程序的识别正确率..................................................34

结论.........................................................................40

致谢.........................................................................41

参考文献.....................................................................42

附录1.......................................................................44

附录2.......................................................................46