【摘要】 提出一种新颖的基于分形理论的苹果属性分类识别方法。通过机器视觉获取苹果彩色图像,进行处理,得到反映苹果色度属性的灰度图象;运用分形理论,通过苹果灰度图象创建灰度特性曲线,对该曲线的分形问题进行研究,得到曲线的分形维数;根据特征维数,建立起苹果属性类别和分形维数间的关系,实现了苹果属性的准确检测和分类。实验结果表明,用分形理论方法进行苹果分类识别,具有良好的抗噪、容错能力,并能有效地克服传统图像分割识别方法适应性差的缺点,提高了分类判别准确率和分类精度。更多还原

【Abstract】 A novel approach is described in this article on apple classes identification based on fractal theory.Machine vision is used to acquire the image of the apple,to pre-process the image information and get the gray image.Based on the gray apple image,you can construct gray image property curve;Using the fractal theory to analyze the fractal prop-erty of the gray image curve,people can get the fractal dimension of the property.The relation between the gray curve fractal dimension and the apple classes is founded.Experiment indicates that this approach is useful in apple classes and owns outstanding virtue in treating complex natural property.更多还原