基于PCNN的多尺度对比度塔图像融合算法

基于脉冲耦合神经网络(PCNN)的基本原理,提出了一种新型的图像融合算法。新算法在对源图像进行多尺度对比度金字塔分解的基础上,将多尺度对比度金字塔作为PCNN的输入,利用PCNN的全局耦合特性和脉冲同步特性进行对比度选择以实现图像融合。新算法利用了源图像的全局特征,符合人的视觉神经系统的生理学特性,实验结果表明了新型融合算法的有效性。     

基于PCNN和最优化评价准则的曲波域图像融合

利用曲波变换能够准确捕获图像高维奇异信息的特点,提出了一种在曲波域中基于脉冲耦合神经网络和最优化评价准则的图像融合方法。该方法用曲波变换对输入图像进行多尺度分解,再利用脉冲耦合神经网络的全局耦合特性对高频子带曲波系数进行选取,定义图像融合的目标函数,根据最优化目标函数确定低频曲波系数的融合权值,进行曲波逆变换得到融合图像。实验结果以及与其他算法的比较分析表明了算法的有效性和优越性。     

基于非下采样contourlet变换与脉冲耦合神经网络的图像融合算法

提出一种基于非下采样contourlet变换(NSCT)与脉冲耦合神经网络(PCNN)的图像融合算法.该方法首先利用非下采样eontourlet变换对输入图像进行多尺度分解、多方向稀疏分解,准确捕获图像中的高维奇异信息,然后利用脉冲耦合神经网络的同步激发特性确定融合规则,选取融合系数,提高融合性能.实验结果表明,算法比小波变换、contourlet变换有更好的融合性能.     

基于区域相关性的高分辨率遥感图像融合算法

针对基于现有的融合框架模型的遥感图像融合算法难以较好地同时保持融合结果的光谱特性和空间分辨率的问题,在分量替换融合框架模型的基础上,提出一种基于图像区域相关性的高分辨率遥感图像融合算法.该算法利用改进的脉冲耦合神经网络模型,根据全色图像和多光谱图像间的局部相关性差异对各多光谱图像进行区域分割,根据区域的相关性的差异确定相应的融合规则.实验结果表明,文中算法能取得很好的融合效果.     

脉冲耦合神经网络自适应图像融合算法研究

作为一种新型的神经网络模型,脉冲耦合神经网络（PCNN）已经在众多领域得到了应用。针对现有脉冲耦合神经网络图像融合算法存在的不足,提出了一种新的自适应PCNN图像融合算法。提取原始待融合图像的互补特征作为PCNN的外部输入,并通过提取待融合图像的对比度特征自适应确定PCNN的链接强度参数;分析了传统PCNN获取最优图像融合结果的方法,探索性地将结构相似度引入到PCNN融合结果的评价中,为PCNN最优融合结果的获取提供了很好的借鉴作用。通过红外和可见光等图像的仿真实验结果表明,提出的融合算法是有效的。     

基于脉冲耦合神经网络与欧拉数的图像检索

在分析基于脉冲耦合神经网络时间序列的图像检索算法的基础上,提出一种新的基于脉冲耦合神经网络的图像检索方法。脉冲耦合神经网络是第三代人工神经网络,能够很好地将二维图像的特征提取成一维的矢量值,而且利用脉冲耦合神经网络提取图像特征时具有平移、旋转、尺度和扭曲不变性。新算法针对灰度图像,利用脉冲耦合神经网络模型对图像进行分解,从而生成与原图像相关的二值图像序列,然后针对二值图像序列中的每一幅二值图像,计算反映其边缘信息的欧拉数,由此构造一个一维的特征矢量。在进行图像检索的时候,使用欧式距离进行图像的相似度度量。新算法具有计算简单、数据量小的特点。实验结果表明,新算法具有较强的鲁棒性和检索精度。     

基于非下采样Shearlet变换与聚焦区域检测的多聚焦图像融合算法

为了提高基于多尺度变换的多聚焦图像融合中聚焦区域的准确性,提出了一种基于非下采样Shearlet变换(NSST)与聚焦区域检测的多聚焦图像融合算法。首先,通过基于非下采样Shearlet变换的融合方法得到初始融合图像;其次,将初始融合图像与源多聚焦图像作比较,得到初始聚焦区域;接着,利用形态学开闭运算对初始聚焦区域进行修正;最后,在修正的聚焦区域上通过改进的脉冲耦合神经网络(IPCNN)获得融合图像。与经典的基于小波变换、Shearlet变换的融合方法以及当前流行的基于NSST和脉冲耦合神经网络(PCNN)的融合方法相比,所提算法在客观评价指标互信息(MI)、空间频率和转移的边缘信息上均有明显的提高。实验结果表明,所提出的算法能更准确地识别出源图像中的聚焦区域,能从源图像中提取出更多的清晰信息到融合图像。     

非下采样Contourlet变换与脉冲耦合神经网络相结合的SAR与多光谱图像融合

由于获取地物波谱信息的波段范围及成像方式的不同,SAR与多光谱图像所得到的信息有很大差异,而且SAR图像会受到严重的相干斑噪声干扰,因此SAR与多光谱图像的融合很难获得满意的效果。考虑到非下采样Contourlet变换(NSCT)相比于其他多尺度几何分析方法的优势,提出了一种NSCT与脉冲耦合神经网络(PCNN)相结合的SAR与多光谱图像融合方法。源图像首先经过NSCT分解获得不同尺度多个方向下的分解系数,将分解系数的高斯拉普拉斯算子能量作为脉冲耦合神经网络模型的输入,具有较大点火频率的系数将被选择作为融合图像的系数,最后经过NSCT重构得到最终的融合图像。实验结果表明,这种算法无论在主观视觉还是在客观指标上都要优于之前的许多算法。     

基于生物视觉特性的红外与可见光图像融合方法

脉冲耦合神经网络PCNN以其在图像分割、目标识别等领域的独特优势而成为当前的研究热点。本文对其在红外与可见光图像融合领域的应用进行了研究,并针对传统脉冲耦合神经网络参数无法自动设定的难题,提出了基于修正PCNN的参数自动设定方案。针对可见光与红外图像融合的大量实验结果表明,本文方法无论在主观视觉效果还是客观评价参数上均明显优于基于多分辨分析的融合算法,对于拓宽PCNN的应用领域有一定价值。     

基于波原子变换与脉冲耦合神经网络的图像融合

针对同一场景多聚焦图像的融合问题,提出一种基于波原子变换和脉冲耦合神经网络PCNN(Pulse Coupled Neural Network)的图像融合算法。首先将待融合图像分别进行波原子变换;其次使用多通道PCNN模型对子带图像进行非线性融合;最后对融合处理的系数进行波原子逆变换得到融合图像。仿真实验结果显示该算法能有效地提取待融合图像的特征信息,在主观视觉效果与客观性能指标上均优于传统图像融合方法。     

Multi-focus image fusion using PCNN

This paper proposes a new method for multi-focus image fusion based on dual-channel pulse coupled neural networks (dual-channel PCNN). Compared with previous methods, our method does not decompose the input source images and need not employ more PCNNs or other algorithms such as DWT. This method employs the dual-channel PCNN to implement multi-focus image fusion. Two parallel source images are directly input into PCNN. Meanwhile focus measure is carried out for source images. According to results of focus measure, weighted coefficients are automatically adjusted. The rule of auto-adjusting depends on the specific transformation. Input images are combined in the dual-channel PCNN. Four group experiments are designed to testify the performance of the proposed method. Several existing methods are compared with our method. Experimental results show our presented method outperforms existing methods, in both visual effect and objective evaluation criteria. Finally, some practical applications are given further.     

用于非高斯系统降维的最小残差熵主元网络

基于最小均方误差的主元分析和主元神经网络是有效的多变量降维统计技术,它们所提取的主元含有系统最大方差.非高斯随机系统的近似模型应当含有系统最大信息熵,但包含最大方差并不一定包含最大信息熵.该文提出一种以最小残差熵为通用指标的非线性主元神经网络模型,并给出了一种基于Parzen窗口密度函数估计的熵近似计算方法和网络学习算法.然后从信息论角度分析了,在高斯随机系统中基于最小残差熵和最小均方差为指标的主元网络学习结果具有一致性.最后以仿真验证该方法的有效性,并与基于最小均方误差的主元分析和主元神经网络方法的计算结果进行对比性分析.     

Medical image registration based on self-adapting pulse-coupled neural networks and mutual information

Medical image registration plays a dominant role in medical image analysis and clinical research. In this paper, we present a new coarse-to-fine method based on pulse-coupled neural networks (PCNNs) and mutual information (MI). In the coarse-registration process, we use the PCNN-clusters’ invariant characteristics of translation, rotation and distortion to get the coarse parameters. And the parameters of the PCNN model are optimized by ant colony optimization algorithm. In the fine-registration process, the coarse parameters provide a near-optimal initial solution. Based on this, the fine-tuning process is implemented by mutual information using the particle swarm optimization algorithm to search the optimal parameters. For the purpose of proving the proposed method can deal with medical image registration automatically, the experiments are carried out on MR and CT images. The comparative experiments on MI-based and SIFT-based methods for medical image registration show that the proposed method achieves higher performance in accuracy.     

Binary Image Thinning Using Autowaves Generated by PCNN

This paper proposes a novel binary image thinning algorithm by using the autowaves generated by Pulse Coupled Neural Network (PCNN). Once the autowaves travelling in different directions meet, the PCNN delivers the thinning results. Four meeting conditions are given for autowaves meeting. If a neuron satisfies one of the four conditions, the pixel corresponding to this neuron belongs to the thinning result. Moreover, the specification of the PCNNs parameters is given, which makes the implementation of the proposed thinning algorithm easy. Experimental results show that the proposed algorithm is efficient in extracting the skeleton of images (such as Chinese characters, alphabet letters, numbers, fingerprints, etc.). Finally, a rate called “R _MSkel” is given to evaluate the performance of different thinning algorithms, and comparisons show that the proposed algorithm has higher “R _MSkel” and costs less time.     

一种新的基于PCNN的图像脉冲噪声滤波算法

为了更好地滤除图像中的脉冲噪声,在分析脉冲耦合神经网络(PCNN)的工作机理的基础上,在PCNN中引入了点火映射图的概念,并提出了一种新的基于PCNN点火映射图的图像脉冲噪声滤波器。该方法首先通过运行PCNN把噪声图像转化为点火映射图,然后利用点火映射图对图像中的脉冲噪声进行定位,最后仅对定位的噪声进行自适应滤波。计算机仿真实验表明,该方法的去噪效果优于传统方法,从而验证了该方法的有效性、合理性。     

一种改进的巴斯分形变换

分析了用多巴斯分形变换进行编码时存在的两个问题，提出了一种改进的分形变换。模拟结果表明其分形系数的动态范围小，分布集中，因此可以提高压缩性能，同时其解码所需的迭代次数少，典型地只需３次。     

Efficient search algorithm on compact S-trees

Based on the S-tree, the modified S-tree is a newly proposed spatial data structure for representing digital binary images, which can support fast search and query in pictorial database. In this paper, we first present a new spatial data structure called the compact S-tree to represent binary images, which uses less memory when compared to the modified S-tree. We then present a fast search algorithm on the compact S-tree. Further, the application to neighbor finding is also investigated. Experimental results show that the proposed search algorithm is faster than the previous known results on the S-tree and the modified S-tree.     

Pulse-coupled neural networks for contour and motion matchings

Two neural networks based on temporal coding are proposed in this paper to perform contour and motion matchings. Both of the proposed networks are three-dimensional (3D) pulse-coupled neural networks (PCNNs). They are composed of simplified Eckhorn neurons and mimic the structure of the primary visual cortex. The PCNN for contour matching can segment from the background the object with a particular contour, which has been stored as prior knowledge and controls the network activity in the form of spike series; The PCNN for motion matching not only detects the motion in the visual field, but also extracts the object moving in an arbitrarily specified direction. The basic idea of these two models is to encode information into the timing of spikes and later to decode this information through coincidence detectors and synapse delays to realize the knowledge-controlled object matchings. The simulation results demonstrate that the temporal coding and the decoding mechanisms are powerful enough to perform the contour and motion matchings.     

Output-threshold coupled neural network for solving the shortest path problems

This paper presents a coupled neural network, called output-threshold coupled neural network (OTCNN), which can mimic the autowaves in the present pulsed coupled neural networks (PCNNs), by the construction of mutual coupling between neuron outputs and the threshold of a neuron. Based on its autowaves, this paper presents a method for finding the shortest path in shortest time with OTCNNs. The method presented here features much fewer neurons needed, simplicity of the structure of the neurons and the networks, and large scale of parallel computation. It is shown that OTCNN is very effective in finding the shortest paths from a single start node to multiple destination nodes for asymmetric weighted graph, with a number of iterations proportional only to the length of the shortest paths, but independent of the complexity of the graph and the total number of existing paths in the graph. Finally, examples for finding the shortest path are presented.     

基于简化血流图小波包域DCT系数决策融合的红外人脸识别

本文提出了一种基于简化血流图小波包域DCT系数融合的红外人脸识别方法。首先,基于人体的皮肤温度分布和温度调节机理,结合红外成像原理及生物传热学知识对人脸的血流模型进行简化,把红外人脸温谱图转换成简化血流图,然后将人脸简化血流图进行三级小波包分解,得到小波包分解树,选取其中识别率最高的若干个节点分别进行DCT变换,得到每个节点的特征矩阵,再通过欧氏距离和三阶近邻分类器得到各选中节点的识别结果,最后将这些结果进行决策融合,得到最终的识别结果。实验结果表明,对血流模型的简化可以在几乎不降低识别的同时,减小时间的复杂度,而在小波包域进行DCT系数融合的方法能提取更加有效的人脸特征,从而提高了红外人脸识别的性能。