A recurrent perceptron learning algorithm for cellular neural networks

A supervised learning algorithm for obtaining the template coefficients in completely stable Cellular Neural Networks (CNNs) is analysed in the paper. The considered algorithm resembles the well-known perceptron learning algorithm and hence called as Recurrent Perceptron Learning Algorithm (RPLA) when applied to a dynamical network. The RPLA learns pointwise defined algebraic mappings from initial-state and input spaces into steady-state output space; despite learning whole trajectories through desired equilibrium points. The RPLA has been used for training CNNs to perform some image processing tasks and found to be successful in binary image processing. The edge detection templates found by RPLA have performances comparable to those of Canny's edge detector for binary images.     

干涉图特征信息自动采集方法

介绍了一种用于计算全息数字波面干涉仪中采用数字图像处理技术采集干涉图的特征信息的方法。它分为图像二值化、条纹细化、条纹修补、跟踪、标记、采样等步骤,对于没有闭合条纹的简单干涉图和有闭合条纹的复杂干涉图,都能得到较好结果。     

基于图像块能量梯度的多聚焦图像融合

针对多聚焦图像,提出一种基于图像分块的融合方法。将源图像分为大小相同数量相等的子块,采用能量梯度算子作为对焦评价函数,计算各个图像子块能量梯度匹配度,设置匹配度阈值分离出源图像中的清晰区域。源图像中的清晰区域直接作为融合图像相应的区域,其它区域的处理中,构造与相应子块能量梯度大小相关的图像序列,以及像素点到各个子块中心距离相关的融合函数,然后用融合函数对图像序列融合。实验结果表明该方法有效性和合理性。     

Analysis of HRTEM images for carbon nanostructure quantification

Image processing algorithms have been developed to extract fringe length, tortuosity and separation from high resolution transmission electron microscopy images. To validate the separation algorithm, a comparison is made between the image-based fringe separation and that obtained by analysis of X-ray diffraction data for a progressively heat-treated carbon black. Agreement is favorable. To illustrate the utility of the analysis parameters for a range of carbon nanostructures, analysis is applied to a series of pyrolytically prepared carbon soots – qualitatively described as containing amorphous, graphitic or fullerenic nanostructure. For all processing, the intermediate image, in the form of a skeletonized binary image of the original high resolution transmission electron micrograph, is shown and found to accurately reflect the nanostructural organization within the carbon as visually observed. Statistical results for each analysis parameter, extracted from the binary images, are presented in the form of histograms and quantitatively distinguish the different carbon nanostructures.     

One-operation image algebra for optical processing

A binary image algebra with one operation of a logic operation followed by a dilation and its modification with three additional operations for threshold processing of grey level images are developed in this paper. All the binary image processing functions and various grey level image processing functions related to morphological operations can be represented by the algebraic structure. The algebra is particularly suitable for parallel processing by optics in a cellular logic image processor architecture.     

IKWI-net: A cross-domain convolutional neural network for undersampled magnetic resonance image reconstruction

Magnetic resonance imaging (MRI) is widely used to get the information of anatomical structure and physiological function with the advantages of high resolution and non-invasive scanning. But the long acquisition time limits its application. To reduce the time consumption of MRI, compressed sensing (CS) theory has been proposed to reconstruct MRI images from undersampled k-space data. But conventional CS methods mostly use iterative methods that take lots of time. Recently, deep learning methods are proposed to achieve faster reconstruction, but most of them only pay attention to a single domain, such as the image domain or k-space. To take advantage of the feature representation in different domains, we propose a cross-domain method based on deep learning, which first uses convolutional neural networks (CNNs) in the image domain, k-space and wavelet domain simultaneously. The combined order of the three domains is also first studied in this work, which has a significant effect on reconstruction. The proposed IKWI-net achieves the best performance in various combinations, which utilizes CNNs in the image domain, k-space, wavelet domain and image domain sequentially. Compared with several deep learning methods, experiments show it also achieves mean improvements of 0.91 dB in peak signal-to-noise ratio (PSNR) and 0.005 in structural similarity (SSIM).     

The parallel processor for the character extraction of digitizing image

ＴｈｅｐａｒａｌｌｅｌｐｒｏｃｅｓｓｏｒｆｏｒｔｈｅｃｈａｒａｃｔｅｒｅｘｔｒａｃｔｉｏｎｏｆｄｉｇｉｔｉｚｉｎｇｉｍａｇｅＨＵＡＮＧＧｕｏｌｉａｎｇ；ＷＵＭｉｎｘｉａｎ；ＪＩＮＧｕｏｆａｎ；ＹＡＯＺｈｏｎｇｂｉｎｇ；ＹＡＮＹｉｎｇｂａｉ（Ｄｅｐｔ．...     

Dispersive correction for the second- and the third-order nonlinear terms of the electrical energy density for a light-dielectric system

Optical and Quantum Electronics - Classification of hyperspectral image (HSI) is widely used for the study of remotely sensed images. Convolutional Neural Networks (CNNs) are one of the most...     

快速消除车辆阴影的多阈值图像分割法

 视频图像分割时的运动阴影由于与被测对象的相似性而常被误判为被测对象,传统的阴影检测方法一般难以满足实时智能交通系统对处理速度的要求。为此,提出用多阈值法分割图像,将灰度化的当前图像与背景差分,再用正、负两个阈值对其二值化,在分割出深色和浅色被测对象的同时消除阴影。实验表明,将这种方法应用于运动车辆的检测,具有速度快和阴影消除效果好等特点,可应用于实时的运动目标检测和跟踪等领域。     

基于Harris角点检测器的指纹细节点提取算法

提出了一种基于Harris角点检测器的指纹细节点提取算法。先将Harris角点检测器应用于增强后的指纹图像,检测出指纹的细节点和曲率变化大的点,随后进行后处理操作。在后处理操作中,根据细节点的空间分布特征,删除虚假点。使用细节点的邻域灰度信息,判定细节点的类型。利用细节点的原始方向以及得到的类型,确定细节点的精确方向。与经典的细节点提取算法相比,不需要对指纹图像进行二值化、细化,直接在灰度指纹图像当中提取细节点,大大减少了计算时间,有效地提高了效率。使用FVC2002指纹数据库测试,结果表明,该算法可靠、快速,适合实际应用。     

基于边缘几何特征和频域相关技术的图像匹配方法

提出了一种基于图像特征提取和变换域相关的图像匹配方法。借助提取图像的大边缘信息，以其几何特征构造边缘方向角曲线作为图像匹配的依据。通过对差分的计算，并比较两幅图像边缘方向曲线的相对链码，得到2幅图像之间可能的旋转角度。对旋转矫正后的图像进行2维傅里叶变换，计算频域内2图的相位相关矩阵，并从中找到相关系数最大的位置，这个位置就代表了2图的坐标位移量，通过平移达到图像间的精确匹配。实验表明：该方法匹配准确，速度快，减少了人为因素的干扰，对于图像间存在较小灰度差和噪声情况均有很好的匹配效果。     

基于灰度图差分梯度的雾化角测量方法北大核心CSCD

雾场边界及雾化角作为雾场的重要特性参数,主要通过图像法进行测量。在图像处理过程中,一般是将灰度图转化为二值化图像,然后依次针对二值化图像进行处理和计算。由于雾场的多相流特性,得到的二值化阈值和图像与实际雾场是否一致缺少评判依据。提出根据喷雾的灰度图像直接处理,得到掩模板并作用于灰度图像,采用图像形态学和迭代方法,计算灰度图像的梯度值。通过得到梯度值最大时的灰度图像,计算雾场边界和雾化角。实验表明,该方法提供了一种雾场边界的数值判断依据,通过梯度最大值判断并提取雾场边界,从而通过程序自动实现雾场边界提取与雾化角拟合测量。     

Soft Compression for Lossless Image Coding Based on Shape Recognition

Soft compression is a lossless image compression method that is committed to eliminating coding redundancy and spatial redundancy simultaneously. To do so, it adopts shapes to encode an image. In this paper, we propose a compressible indicator function with regard to images, which gives a threshold of the average number of bits required to represent a location and can be used for illustrating the working principle. We investigate and analyze soft compression for binary image, gray image and multi-component image with specific algorithms and compressible indicator value. In terms of compression ratio, the soft compression algorithm outperforms the popular classical standards PNG and JPEG2000 in lossless image compression. It is expected that the bandwidth and storage space needed when transmitting and storing the same kind of images (such as medical images) can be greatly reduced with applying soft compression.     

高功率激光器自动准直系统图像处理的可信度评估

自动准直系统是将采集的图像进行特定算法处理获取光束的位置信息,驱动电机调整光斑到指定的位置,其中图像质量对光束定位的精确性影响很大,为尽量避免受严重噪音和光束畸变影响的图像进入图像处理流程而产生准直结果的错误判断,保证自动准直系统图像处理结果的精确性,试图提供一种较为简单的判断依据,采用基于Monte-Carlo模拟方法,建立了远场准直过程中三种主要噪音的干扰评估模型,以测量不确定度作为定量判断处理图像与否的特征参数,将该结果做成统计图表作为判断依据。结果表明,在设定合理测量不确定度阈值的情况下,该特征参数能够较好地排除不良图像,从而提高准直结果的可信度。     

Double image encryption based on phase-amplitude hybrid encoding and iterative phase encoding in fractional Fourier transform domains

A new method for double image encryption is proposed that is based on amplitude-phase hybrid encoding and iterative random phase encoding in fractional Fourier transform (FrFT) domains. In the iterative random phase encoding operation, a binary random matrix is defined to encode two original images to a single complex-valued image, which is then converted into a stationary white noise image by the iterative phase encoding with FrFTs. Compared with the previous schemes that uses fully phase encoding, the proposed method reduces the difference between two original images in key space and sensitivity to the FrFT orders. The primitive images can be retrieved exactly by applying correct keys with initial conditions of chaotic system, the pixel scrambling operation and the FrFT orders. Computer simulations demonstrate that the encryption method has impressively high security level and certain robustness against data loss and noise interference.     

自适应参考图像的可见光与热红外彩色图像融合算法

可见光与红外热图像的彩色图像融合技术是现今国内外高性能夜视技术发展的重要方向之一,该技术有效提高了人们对目标的探测和场景理解能力。目前常用的色彩传递算法多属于基于单幅参考图像的全局色彩传递算法,彩色融合图像的色调受到参考图像的影响较大,在实际应用中难以保证对各类场景的适应性。针对常规YUV空间色彩传递彩色图像融合算法的环境适应性问题,通过对植物、城镇和海天三类典型场景的分类与统计,发现了典型场景在UV通道的均值和标准差具有的较为明显的分类特性,由此提出了一种基于UV通道均值和标准差的自适应参考图像构造方法,使得可见光与热红外彩色图像融合算法具有较常规算法更好的环境适应性,融合图像的色彩具有较好的自然感,且算法处理量较小,对现有实时硬件融合处理算法的运算速度影响不大,是一种环境适应性强的自然感彩色融合处理算法。     

基于灰度相关的红外隐身效果评价方法研究

提出了一种基于图像灰度相关算法的红外隐身效果评价方法.通过对采用红外隐身措施前后目标图像的相关处理,提取相关峰的峰显著系数、峰.相关能量比和峰-背景比这三个特征参量来定义红外隐身概率作为衡量红外隐身效果的客观评价指标.采用人眼对红外热图进行判识的统计结果来确定红外隐身概率中各参量的统计权重,从而使计算机图像处理方法所得结果与实际人眼的主观判识结果一致.实验结果表明:此方法适用于对地面典型军事目标的红外隐身效果进行客观评价.     

Design for CNN Templates with Performance of Global Connectivity Detection

The cellular neural/nonlinear network (CNN) is a powerful tool for image and video signal processing, robotic and biological visions. This paper discusses a general method for designing template of the global connectivity detection (GCD) CNN, which provides parameter inequalities for determining parameter intervals for implementing the corresponding functions. The GCD CNN has stronger ability and faster rate for determining global connectivity in binary patterns than the GCD CNN proposed by Zarandy. An example for detecting the connectivity in complex patterns is given.     

Grayscale-Memory Spatial Light Modulator with Polymer-Dispersed Phase-Transition Liquid Crystal

A new storage-type spatial light modulator has been developed which uses a polymer-dispersed phase-transition liquid crystal film and a photoconductor for high-brightness projection displays. The liquid crystal film containing micrometer-sized chiral nematic liquid crystal domains subdivided by a unique honeycomb polymer network functions as a memory layer for an input image. The liquid crystal domains independently maintain the binary alignment states in the opaque (light-scattering) cholesteric phase or transparent nematic phase. Grayscale images can be displayed according to the spatial distribution of the bistable liquid crystal domains. The device exhibits rapid response of a few ms and a limiting resolution of 32 lp/mm. It has the advantages of high transmittance without polarizers and ease of fabrication due to the free-standing liquid crystal film.     

Look-up table and Gaussian filter-based inverse halftoning method excellent in gray-scale reproducibility of details and flat regions

Halftoning is a technique in which gray-scale images are converted into binary ones, and has been used widely in many practical image processing applications. On the other hand, inverse halftoning is a technique of reproducing a grayscale image from a binary one. One of the sophisticated inverse halftoning methods which can yield a high quality grayscale image is that based on a look-up table (LUT). In this paper, a new inverse halftoning method based on a combination of this LUT method and a Gaussian filter is discussed. Here, the output gray-scale image is reproduced by blending the output of the LUT-based inverse halftoning and that of the Gaussian filtering appropriately in accordance with local statistics of the image. The effectiveness and the validity of the described method are verified through experiments.