精品毕业论文--基于模式识别的水果智能分类系统设计.docVIP

基于模式识别的水果智能分类系统 摘 要 本论文综合运用了数字图像处理，模式识别的理论来构建起一个简单的水果智能分类系统。实现了在相同条件下拍摄的水果图片的特征提取和种类识别，在此基础上设计出了基于人工神经网络的水果智能分类器，由计算机自动调整神经网络中各个权值，达到水果种类识别的自动化。 数字图像处理对源位图进行了加工，是特征提取的基础。数字图像处理的理论涉及到彩色图像的灰度化、中值滤波、二值化、轮廓提取、种子填充、轮廓跟踪等。其中，二值化采用了基本自适应门限的方法。 模式识别包括了特征提取和分类器的设计，是种类识别的关键。特征提取主要利用了水果的几何特征，反映了水果的大小和形状。分类器的设计主要采用了人工神经网络的方式来实现，具体说来是利用了神经网络中反向传播算法来进行网络训练，并利用训练结果完成了水果种类的智能识别。 关键词：特征提取 人工神经网络 二值化 基本自适应门限 反向传播算法 A Intellective System for Fruit Classification Based on Pattern Recognition Abstract In this paper, we apply the theory of digital image processing and pattern recognition to construct a simply and intellective system for fruit classification based on pattern recognition. We have already fulfilled characteristic withdrew and type recognition for the pictures of fruit which are photographed under the same condition .We have also designed a categorize machine based on artificial neuro-network , which can adjust the weights of neuro-network automatically by computer in order to recognize the type of the fruit. Digital image processing deals with the original bitmap ,which is the basis of characteristic withdrew .The theory of digital image processing refers to the gradation of color image ,median filter ,image binary, outline withdrew ,the seed fills ,outline track and so on. Among them, image binary makes use of the basic auto-adapted threshold method. Pattern recognition involves characteristic withdrew and the design of categorize machine, which are the keys of type recognition. The characteristic withdrew has mainly used fruit's geometry characteristics ,which reflect fruit’s size and shape .The categorize machine is designed by means of artificial neuro-network, which uses the algorithm of Back-Propogation in detail and completes the fruit type intelligent recognition by using the training results. Keywords: characteristic withdrew, artificial neuro-network, image binary, basic auto-adapted threshold, the algorithm of Back-Propogation. 目录 摘要 I Abstract II 第1章 绪论 1 1.1