纹理细节恢复的图像超分辨率重建算法

针对现有单图像超分辨率方法在重建过程中容易忽略原图像中不同结构-纹理的差异与联系，导致生成的高分辨率图像缺乏纹理细节并存在伪影的问题，提出了纹理细节恢复的图像超分辨率重建算法。该方法由梯度分支、纹理分支和图像超分辨率分支组成。其中，在梯度分支和纹理分支之间使用了类注意力模块处理二者的特征混淆问题，并通过双向特征融合模块实现了对结构特征与纹理特征的相互促进，作为先验信息以达到纹理细节信息增强的目的。此外，在图像超分辨率分支还通过构建特征恢复模块，利用浅层和深层信息帮助网络保留了图像中更丰富的上下文信息和纹理细节。该方法通过在DIV2K数据集上进行了网络训练，并在5个基准测试集Set5、Set14、BSD100、Urban100和MANGA109上进行了实验，峰值信噪比(PSNR, Peak Signal to Noise Ratio)分别：37.88dB、33.28dB、32.0781dB、31.89dB、38.39dB,相比现有方法均有显著提升。实验结果表明，本文方法获得了有效的重建图像并且保留更多的图像细节，生成具有边缘清晰和逼真细节的超分辨率图像。     

基于模糊推理的纹理图像融合方法研究

研究了模糊逻辑方法在图像特征层融合中的应用,介绍了两种模糊推理方法—Mamdani型和Sugeno型模糊逻辑推理系统,以及基于Sugeno型模糊推理的自适应模糊神经网络系统;在红外-可见光双通道图像融合实验中,采用Gabor滤波提取两幅图像的纹理特征,并分别采用两种模糊推理方法进行特征层融合。对融合效果进行了分析,结果表明两种模糊逻辑方法在不同应用环境下都可以取得了较好的融合效果。     

三维Gabor滤波器与支持向量机的高光谱遥感图像分类

根据高光谱遥感图像的特点及二维Gabor滤波器纹理分割的原理,提出了一种基于三维Gabor滤波器的高光谱遥感图像分类方法。三维Gabor滤波器能够对高光谱遥感图像所有波段同时进行滤波,将大量的图像信息抽取为少量的不同尺寸、方向和波谱的响应,极大减少了高光谱遥感图像纹理信息提取的计算量。利用不同方向和尺寸的三维Gabor滤波器对祁连山黑河流域上游地区的Hyperion影像全波段进行滤波处理,获取26个纹理响应特征,并分析不同纹理对不同地物的区分度。利用自动子空间划分的波段指数(BI)进行波段选择,选取不同的波段组合进行试验,寻找最佳降维幅度。按照纹理对不同地物响应的区分度逐一加入三维Gabor纹理特征,利用三维Gabor纹理辅助光谱信息,运用支持向量机(SVM)的方法进行监督分类。结果表明,基于三维Gabor纹理和自动子空间BI波段选择的SVM分类方法能够在有效降低光谱维数的同时,提高高光谱遥感图像分类的精度和效率。     

暗通道与交叉通道多先验联合多光谱超分辨率算法

针对多光谱遥感图像超分辨率重建易受噪声和色差影响的问题,提出了暗通道与交叉通道多先验联合的多光谱超分辨率算法。首先,在传统全变分先验的基础上,引入了暗通道先验与交叉通道先验。然后,基于最大后验概率估计理论,建立了多先验联合的多光谱超分辨率重建算法。所提算法可实现图像边缘信息恢复、图像纹理信息恢复、噪声抑制、阶梯效应抑制和色差抑制,进而综合提升了重建图像的质量。最后,开展了实验验证,结果表明,在不同信噪比(10～40 dB)和色差干扰下,相比已有算法,所提算法的重建效果显著提升。     

点扩散函数高斯拟合估计与遥感图像恢复

为了减轻或消除航天遥感相机成像过程中图像退化造成的模糊,突出图像的特征目标,对获取的图像进行了恢复处理。首先,采用陷波滤波器在频率域对遥感图像进行了去噪预处理。然后,通过图像中具有刀刃边缘的地物估计成像系统的退化函数,即点扩散函数;同时,利用高斯拟合对估计的点扩散函数进行校正。最后,利用拟合后的点扩散函数,采用自适应维纳滤波对图像进行恢复。实验结果表明：陷波滤波器基本消除了图像中叠加的条带噪声。与原图相比,细节图像恢复后其方差增大4.395,灰度平均梯度增大1.799,Laplacian梯度增大10.014,图像目视效果更清晰。高斯拟合的点扩散函数用于遥感图像恢复,减轻了图像模糊,使图像细节突出,纹理清晰,利于图像的判读和分析。     

双密度复小波与系数相关性的红外图像去噪北大核心CSCD

为了去除红外图像中高斯噪声的同时,更好地保持和恢复图像的纹理边缘和细节特征,提出了基于双密度复小波与系数相关性的红外图像去噪方法。该方法充分利用双密度双树复小波在图像处理上的优势:图像信息的平移不变性、图像纹理细节的多方向选择性等,基于对图像小波系数分布的假设,根据当前小波系数与其父、子小波系数的相关性,对无噪的小波系数作贝叶斯估计,以恢复无噪的红外图像,最后对去噪图像进行引导滤波,去除图像的波纹效果。实验数据显示,该方法在EPI和FSIM以及图像的视觉效果上优于部分现有算法,证明该方法在噪声去除、纹理边缘的保持和恢复上具有更好的性能。     

基于判别字典学习与形态成分分解的多源图像融合

提出了一种基于形态成分分析的多源图像融合方法。为了将源图像中不同形态结构的卡通-纹理成分分离,把图像的分解问题转化为图像的分类问题,设计了卡通纹理判别字典学习模型。考虑到图像分解不仅与字典有关,还与分解的策略有关,设计了一种新的图像分解模型。在模型中,将纹理成分看成叠加在源图像卡通成分上的噪声,引入非局部均值相似性的一致性正则项,来约束稀疏编码系数的解空间。根据对应成分的编码系数l1范数值最大来选取融合图像的编码系数。实验结果表明,无论是在视觉效果还是在客观指标上,方法都具有更好的融合性能。     

复杂背景下钢索图像的纹理分割与边界识别

针对复杂背景下钢索图像难以准确分割的问题,提出一种基于纹理分析的钢索图像分割与边界识别方法.采用基于模糊Hough变换的纹理方向检测方法确定钢索走向,利用边缘方向密度直方图作为纹理特征,对与钢索纹理方向相应的边缘方向赋予不同权重,抑制纹理分割中背景的干扰,对钢丝绳图像进行聚类分割,采用检测平行直线的方法确定其边界,并根据算法参量对边界进行修正.在实验中,对比了边缘方向密度直方图特征与灰度共生矩阵、局部二值模式在钢索图像纹理分割中的结果与计算时间,结果表明边缘方向密度直方图特征计算速度快、受背景干扰小,分割准确率高.本文方法无须预先训练,受背景干扰小,可以准确地识别出钢索并确定其边界,能满足钢丝绳视觉检测的要求.     

基于纹理显著性的微光图像目标检测

微光图像对比度较低,目标显著性不明显,目标自动探测难度大.针对此问题,本文提出一种噪声鲁棒性较好的图像局部纹理粗糙度算法,并给出一种适用于微光图像显著分析的纹理显著性算法.首先,提出一种新的局部纹理粗糙度算法,该算法利用最佳尺寸计算局部纹理粗糙度,对纹理图像进行加噪实验,与基于局部分形维的粗糙度方法相比,本文局部纹理粗糙度算法表现出较好的噪声鲁棒性;其次,在提取图像粗糙度特征图的基础上,给出一种针对纹理的显著性度量算法;最后,将纹理显著性算法应用于微光图像目标检测,实验结果证明了该算法的有效性.     

基于字典学习的稀疏角度采样光声信号重建

光声成像兼具光学成像的高对比度和超声成像对深层组织的高分辨率等优点,在生物医学成像领域具有巨大的潜力,而且发展十分迅速;光声成像通过在多个角度进行光声信号的采集,可以获得生物组织的二维或三维光学吸收分布图像;但实际的光声成像往往因硬件条件和成像时间的制约而难以采集角度足够多的光声信号;在信号采样不足的情况下,光声图像的重建质量会严重下降,出现大量伪迹。针对该问题,提出了一种基于字典学习与稀疏表示的恢复重建算法,采用该算法对光声信号进行预处理,并进行仿真实验。结果表明:与不经过光声信号超分辨率重建的时间反演法图像重建结果相比,经所提算法处理后的光声重建图像的伪迹显著减少,细节更加清晰,峰值信噪比提高了8 dB左右;不同信噪比下的仿真实验验证了所提出算法具有良好的稳健性。     

Study on application of complementary Golay code into high frame rate ultrasonic imaging system

The high frame rate (HFR) ultrasonic imaging system, which is developed with limited diffraction beams, constructs images at a high frame rate. However the rectangular imaging area, to some extent, restricts the far field imaging information. At the same time, by virtue of one transmission event for constructing image, the system suffers from low SNR. In this paper we present a computationally efficient method to construct sector mode image and to increase the SNR in HFR system. The method uses orthogonal complementary Golay coded excitation to realize two transmit and receive events. Each emission simultaneously transmits two plane waves with different transmission angle. Then according to Golay code orthogonality, the received echo signals related to different angles are isolated and used to construct two images of different imaging area by HFR method. Finally the two images are synthesized to one frame of sector mode image.     

A high resolution ultrasonic microscope

The construction and imaging capabilities of a scanning high resolution ultrasonic microscope are considered in this paper. Theoretical expressions for the resolution limits of the microscope and some of the most important properties of the ultrasonic lenses are given. Ultrasonic through-transmission and back-scattering images of different objects are presented and a quasi-three-dimensional display mode is introduced. A resolution of about 4 μm is demonstrated at 300 MHz by the present ultrasonic microscope.     

一种基于规则带特征的图案布匹检测新方法

为了对具有纹理图案的纺织品缺陷进行准确有效的检测,提出了一种新的基于规则带( regular band,RB)特征的纹理图案布匹瑕疵检测方法.首先利用规则带分析提取图像的各子带纹理特征;然后对各子带图像特征求梯度,并通过融合各子带梯度来获得纹理梯度,使其在纹理梯度中能有效地突出纹理区域的边界;最后在此基础上,结合分水岭...     

On the feasibility of remote palpation using acoustic radiation force

A method of acoustic remote palpation, capable of imaging local variations in the mechanical properties of tissue, is under investigation. In this method, focused ultrasound is used to apply localized (on the order of 2 mm3) radiation force within tissue. and the resulting tissue displacements are mapped using ultrasonic correlation based methods. The tissue displacements are inversely proportional to the stiffness of the tissue, and thus a stiffer region of tissue exhibits smaller displacements than a more compliant region. In this paper, the feasibility of remote palpation is demonstrated experimentally using breast tissue phantoms with spherical lesion inclusions, and in vitro liver samples. A single diagnostic transducer and modified ultrasonic imaging system are used to perform remote palpation. The displacement images are directly correlated to local variations in tissue stiffness with higher contrast than the corresponding B-mode images. Relationships between acoustic beam parameters, lesion characteristics and radiation force induced tissue displacement patterns are investigated and discussed. The results show promise for the clinical implementation of remote palpation.     

MRI texture analysis on texture test objects, normal brain and intracranial tumors

Texture analysis was performed in three different MRI units on T1 and T2-weighted MR images from 10 healthy volunteers and 63 patients with histologically confirmed intracranial tumors. The goal of this study was a multicenter evaluation of the usefulness of this quantitative approach for the characterization of healthy and pathologic human brain tissues (white matter, gray matter, cerebrospinal fluid, tumors and edema). Each selected brain region of interest was characterized with both its mean gray level values and several texture parameters. Multivariate statistical analyses were then applied in order to discriminate each brain tissue type represented by its own set of texture parameters. Texture analysis was previously performed on test objects to evaluate the method dependence on acquisition parameters and consequently the interest of a multicenter evaluation. Even obtained on different sites with their own acquisition routine protocol, MR brain images contain textural features that can reveal discriminant factors for tissue classification and image segmentation. It can also offer additional information in case of undetermined diagnosis or to develop a more accurate tumor grading.     

Image fusion algorithm based on redundant-lifting NSWMDA and adaptive PCNN

Pyramid decomposition in the NSCT transformation is a band-pass filtering process in the frequency domain where different scales of images are orthogonal. However, from the perspective of the image content, correlation is likely to exist between the fused images, and this kind of decomposition makes images of different scales contain redundant information, as a result of which the fused image may not capture the subtle information from the original images. In order to overcome the above-mentioned problem, an effective image fusion method based on redundant-lifting non-separable wavelet multi-directional analysis (NSWMDA) and adaptive pulse coupled neural network (PCNN) has been proposed. The original images are firstly decomposed by using the NSWMDA into several sub-bands in order to retain texture detail and contrast information of the images, and then adaptive PCNN algorithm is applied on the high-frequency directional sub-bands to extract the high-frequency information. The low-frequency sub-bands are evaluated by weighted average based on Gaussian kernel with a chosen maximum fusion rule. Results from experiments show that the proposed method can make the fused image maintains more texture details and contrast information.     

Adaptive diagnostic ultrasonic imaging

Medical ultrasonic imaging systems assume a constant, fixed acoustic propagation velocity in tissue. This assumption allows the focusing of ultrasonic pulses in a simple way with transducer arrays with electronic delay lines. However, soft tissue acoustic velocities actually range from 1350 m/s to 1725 m/s, and the basic focusing procedure fails to obtain high quality images on some patients. Here we describe different adaptive techniques that allow focusing through such inhomogeneous tissues.     

Quantitative characterization and comparative study of feather melanosome internal morphology using surface analysis

A successful feather development implies in a precise orchestration of cells in the follicle, which culminates in one of the most complex epidermal structures in nature. Melanocytes contribute to the final structure by delivering melanosomes to the barb and barbule cells. Disturbance to the tissue during the feather growth can damage the final structure. Here, melanosomes seen in an unusual outgrowth on the barb cortex of a flight feather are reported and compared to commonly observed melanosomes embedded in the cortex. Transmission Electron Microscopy in scanning-transmission mode (STEM) generated images coupled with secondary electron detection. The two classes of melanosomes were registered on images combining transmitted and secondary electron signals. Image processing allowed surface analyses of roughness and texture of the internal morphology of these organelles. Results showed that the two classes of melanosomes are significantly distinct internally, indicating that different physiological processes up to feather maturation could have occurred. Surface analysis methods are not regularly used in cell biology studies, but here it is shown that it has great potential for microscopic image analysis, which could add robust information to studies of cell biology events.     

Ultrasonic attenuation estimation in-vivo using the difference ratio correction method.

In recent years, B scanners have been widely used clinically. Obviously, it is of particular importance to use the commercial B scanners to characterize tissue by estimating its ultrasonic attenuation in vivo. However, there are a lot of difficulties in doing so because the output of a B scanner is affected by many unknown factors. In this paper, a time domain method named the difference ratio correction (DRC) method is proposed to estimate the ultrasonic attenuation of tissue in vivo. In this method, three tissue-mimicking phantoms with known acoustical properties were employed to eliminate the instrumentation errors of the B scanner and the measuring system and to correct beam diffraction for correct attenuation estimation. Other advantages of this method are that it is very convenient to apply this method clinically and there is no need to change the inner construction of the B scanner because this method only utilizes the video output. Experimental and clinical results have proved the validity of this method.