《神经网络算法在医学图像压缩中的应用研究》由会员分享，可在线阅读，更多相关《神经网络算法在医学图像压缩中的应用研究（37页珍藏版）》请在金锄头文库上搜索。
1、I神经网络算法在医学图像压缩中的应用研究神经网络算法在医学图像压缩中的应用研究摘摘 要要随着图像处理技术的不断进步和发展，数字成像技术在医学中的应用日益广泛， 产生了大量的数据，给图像的存储和传输技术提出了严峻的挑战，解决它的关键技术 之一就是医学图像压缩技术。 传统的图像压缩方法很多，但大多存在压缩比低、误差大的缺点。基于神经网络 的图像压缩技术，在理论和技术上开辟了图像压缩的新途径。BP（Back Propagation） 网络是1986年由Rumelhart和McCelland为首的科学家小组提出，是一种按误差逆传播算 法训练的多层前馈网络，是目前应用最广泛的神经网络模型之一。BP网络能
2、学习和存 贮大量的输入-输出模式映射关系，而无需事前揭示描述这种映射关系的数学方程。它 的学习规则是使用最速下降法，通过反向传播来不断调整网络的权值和阈值，使网络 的误差平方和最小。BP神经网络模型拓扑结构包括输入层（input） 、隐层(hide layer)和 输出层(output layer)。BP神经网络算法具有以下优点：网络实质上实现了一个从输入到 输出的映射功能、网络能通过学习带正确答案的实例集自动提取“合理的”求解规则、 网络具有一定的推广能力。 BP神经网络是一种由输入空间到输出空间的数学映射，能以任意精度逼近任何线 性或非线性映射。基于人工神经网络的图像压缩理论与方法成为当今
3、图像处理领域的 研究热点。 【关键词关键词】医学图像压缩；误差逆传播算法；BP神经网络；数学映射；自学习；IIThe Application Of Neural Network Algorithm In Medical Image CompressionABSTRACTAlong with the advance of image processing technology and development of digital imaging technology has been used increasingly in medical, produced a large amount of
4、 data, for image storage and transmission technology put forward the severe challenge, it is one of the key technology of medical image compression technology. The traditional image compression method are many, but most of the low compression ratio, the error big shortcoming. Image compression techn
5、ology based on neural network, on the theory and technology has opened up a new way of image compression. BP (Back Propagation) network is a group of scientists headed by Rumelhart and McCelland in 1986, is a kind of according to the error Back Propagation algorithm training of the multilayer feedfo
6、rward network, is currently one of the most widely used neural network model. BP network can learn and store a lot of input - output model mapping, without prior reveal describe the mathematical equations of the mapping relationship. Its learning rule is to use the steepest descent method, by back p
7、ropagation to constantly adjust the network weights and threshold, minimize the error sum of squares of the network. BP neural network model of topological structures include input layer (input), the hidden layer (hide layer) and output layer (output layer). BP neural network algorithm has the follo
8、wing advantages: essentially implements a mapping from input to output function, the network can take the correct answer by learning instance set automatically extracting rules of “reasonable“ solving, network generalization ability. The BP neural network is a kind of mathematical mapping from input
9、 space to the output space, can approach any linear or nonlinear mapping with any degree of accuracy. Image compression based on artificial neural network theory and method become a research hotspot in the field of image processing. 【Keywords】Medical image compression; Error back propagation algorit
10、hm. The BP neural network; Mathematical mapping; Since the study; III目录目录引言引言.- 1 - 第第 1 章章 绪论绪论.- 2 - 1.1 课题的研究背景及意义.- 2 - 1.2 国内外研究现状.- 4 - 第第 2 章章 图像压缩与神经网络简述图像压缩与神经网络简述.- 6 - 2.1 几种典型的图像压缩编码技术.- 6 - 2.2 图像压缩编码的性能指标.- 7 - 2.3 神经网络模型及应用.- 8 - 第第 3 章章 基于标准基于标准 BP 神神经经网络的图像压缩网络的图像压缩.- 10 - 3.1 BP 算法的基本思想.- 10 - 3.2 BP 算法研究.- 10 - 3.3 BP 学习算法.- 12 - 3.4 基于 BP 算法的神经网络用于图像压缩编码.- 15 - 3.5 基于 BP 网络的图像压缩算法仿真.- 16 - 结论和展望结论和展望.- 21 - 致谢致谢.- 22 - 参考文参考文献献.- 23 - 附录附录 A 外文文献及译文外文文献及译文.
