一种基于Mumford-Shah模型的红外图像边缘检测方法

 Chan-Vese模型是一种优秀的简化Mumford-Shah模型。然而Chan-Vese模型是以两个同质区域为基础建立的，这并不符合红外图像的特点，导致直接应用该模型处理红外图像时可能失败。针对这一问题，提出了一种适用于红外图像边缘检测的改进Mumford-Shah模型，并对该模型中目标边缘的保持、停止准则的建立及算法速度的提高作了详细讨论。实验表明，改进Mumford-Shah模型能够克服Chan-Vese模型在对红外图像边缘检测时不能跨越过渡区域的缺点，有效地检测出目标边缘。     

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。     

Infrared object detection using global and local cues based on LARK

Object detection has become a challenging problem in computer vision. Locally Adaptive Regression Kernel (LARK) based detection methods are able to produce visually pleasing results without any training. We in this paper present an effective object detection method by exploring global and local cues based on LARK features. First, we encode the local context similarity by exploiting region Structural LARK (SLARK) features, which measure the likeness of a pixel to its surroundings in the query image and the test image. Second, a global constraint based on SLARK features via Heat equation is learned to detect similar features in the test image. Results from matrix cosine similarity are computed to estimate similar regions between these computed features. A compactness score is provided to refine these regions. Next, we detect the location of objects in the test image using non-maxima suppression. We show in experiments that the proposed method significantly outperforms other methods on the infrared image datasets, localizing the objects in the test images effectively.     

Healthy and unhealthy red blood cell detection in human blood smears using neural networks

One of the most common diseases that affect human red blood cells (RBCs) is anaemia. To diagnose anaemia, the following methods are typically employed: an identification process that is based on measuring the level of haemoglobin and the classification of RBCs based on a microscopic examination in blood smears. This paper presents a proposed algorithm for detecting and counting three types of anaemia-infected red blood cells in a microscopic coloured image using circular Hough transform and morphological tools. Anaemia cells include sickle, elliptocytosis, microsite cells and cells with unknown shapes. Additionally, the resulting data from the detection process have been analysed by a prevalent data analysis technique: the neural network. The experimental results for this model have demonstrated high accuracy for analysing healthy/unhealthy cells. This algorithm has achieved a maximum detection of approximately 97.8% of all cells in 21 microscopic images. Effectiveness rates of 100%, 98%, 100%, and 99.3% have been achieved using neural networks for sickle cells, elliptocytosis cells, microsite cells and cells with unknown shapes, respectively.     

基于去模糊空间变换RCNN的毫米波图像目标检测（英文）

提出一种包含去模糊的空间变换区域卷积神经网络的目标检测算法.首先,基于主动毫米波圆柱扫描成像原理对人体进行三维成像(频率24~30 GHz),建立毫米波图像数据集.然后,估计毫米波图像的模糊核,通过卷积去噪网络获得图像先验知识,将其集成到半二次分裂的优化方法中,以实现非盲目去模糊.最后,由定位网络、网格生成器和采样网络三部分组成空间变换网络,将它融入到特征提取网络中,在去模糊后实现目标检测.通过该非盲目去模糊算法得到的图像的峰值信噪比可达27.49 dB,目标检测算法的平均精度可达80.9%.实验结果表明,与现有的先进方法相比,该方法可以有效地提高图像质量和检测精度,为毫米波图像中隐藏危险品的目标检测提供了新的技术支持.     

一种基于圆内接直角三角形的圆检测方法

提出一种基于圆内接直角三角形的圆检测方法.创建基于有效像素位置的数组存储空间,将图像空间的搜索任务转移到压缩的数组空间中完成,配以特定的搜索算法,降低了该方法复杂度.在存储空间中分级别搜索直角三角形,将圆参量的求解转化为三角形参量的计算,避免了反复的平方、开方运算,减少了运算时间.在结果计算过程中,将单个直角三角形求解结果进行聚合,最终得到合理全局解.合成图以及实际图像的实验结果表明,该方法用于一般图像的圆检测时,具有独特的抑制噪声以及非圆上边缘点干扰的优势,实验快速、可靠,具有可重复性,检测精度小于1 pixel.     

基于标准互相关函数的图像分辨率自动标定方法

理想规则图像特征经标准互相关函数匹配后,相似函数C*(x)可用确定的解析式Z(x)表达.但是对于含噪声图像,相似函数C(x)较之C*(x)发生了变化,但存在"零相似不变性".将此原理应与于显微视觉中图像物面分辨率的在线标定,推导了矩形图像特征的一维相似函数解析式Z(x);求解C(x)=0作为Z(x)的近似,并给出了具体标定算法;通过仿真图像实验,给出标定算法的正确度在0.1～0.2 pixel;最后,将标定算法应用于可连续变焦的微对准装配系统,实测算法的精密度可达为0.08 pixel.实验结果表明,基于"零相似不变性"的标定算法是具有实用性的亚像素标定方法.～290nm波长激发下,310～390nm范围内产生荧光光谱,峰位波长在350nm,最佳激励波长为250nm;采用垂直偏振片(起偏角为0°)起偏照射样品,发现偏振荧光峰位不变,荧光峰强度随检偏角的增大而呈明显的线性递减关系.由实验数据计算得到荧光偏振度为0.783,表明分子具有一定确定取向;另外,分析认为酸性橙Ⅱ产生荧光,是由于分子中含有苯环和萘环结构吸收紫外光能量,以及氮键在光子作用下形成顺式异构体的激发单线态后,两者发射的光子所致.整个结果对酸性橙Ⅱ在食品中的违禁使用检测、特性表征、以及分子规律的更深入研究,有一定的参考价值.     

Analog Metal-Oxide-Semiconductor Integrated Circuit Implementation of Approach Detection with Simple-Shape Recognition Based on Visual Systems of Lower Animals

We proposed an analog network for motion detection of an approaching object with simple-shape recognition based on the visual systems of lower animals. Locusts can detect an approaching object by performing a simple process in a descending contralateral motion detector (DCMD) existing in their brain, which senses the increase in size and expansion velocity of the image projected on a retina just before collision. The responses correspond to the approaching velocity and direction. Frogs can recognize a simple shape by performing a simple process in a tectum and thalamus existing in their brain based on retinal information; this is called the Ewert von Seelen model. The proposed network was constructed with simple analog metal-oxide-semiconductor (MOS) circuits. Simulation results with a simulation program with integrated circuit emphasis (SPICE) showed that the network was able to detect the approaching velocity and direction, and also recognized simple shapes such as a circle, square, triangle and rectangle.     

An algorithm for object identification based on hyperplane separation but tolerant to illumination and viewing angle variations

A new algorithm for object identification by their images is proposed. It combines tolerance against imaging conditions variations proper to morphological methods for image analysis by Pyt’yev with computational simplicity of methods based on separating classes of images with hyperplane. The algorithm was tested on images taken in different illuminations and views. It can be applied to analyse microscopic image in biophysics, acoustic signals in geophysics, Earth resources using satellite data, etc.     

红外目标湍流退化图像的优化复原算法

提出了基于最速下降法的湍流退化图像盲目复原算法。将图像转换到频域中，建立一个基于目标图像和点扩展函数频谱的目标函数，通过迭代方式采用最速下降优化方法来极小化该目标函数，并利用傅里叶变换和反变换将目标图像和点扩展函数在频域和空域之间进行变换，在每次迭代中交替加入约束条件进行反复修正，以便取得预期的图像恢复效果，增强算法的稳定性和抗噪能力。针对红外目标湍流退化图像，在微机上对算法进行了一系列复原验证实验。实验结果表明：该文算法复原效果稳定，抗噪能力强，具有实用价值。     

半覆盖锥束CT中扁平物体的高效反投影滤波重建

全覆盖圆轨迹扫描的成像视野受探测器宽度限制, 对于大物体的成像效率较低.半覆盖扫描可以将成像视野扩展近1倍, 图像重建首推使用反投影滤波型算法. 反投影滤波型算法按PI线重建, 各PI线积分区间的不一致性导致通信和计算消耗大, 影响重建效率. 针对半覆盖成像中扁平形状物体的重建问题, 提出了一种改进的反投影滤波型算法, 且证明了当扁平物体的厚度小于2Rsin(2π/N_p) (R为扫描半径, N_p 为圆扫描一周均匀采集的投影数量)时, PI线积分区间的不一致性在数值计算过程中的误差是可以忽略的. 改进后的算法相比原半覆盖反投影滤波算法具有两个明显的优势: 一是数值计算过程中角度循环移至PI线循环之外, 算法的通信需求显著降低; 二是投影数据求导、反投影和沿PI线滤波三个步骤均能够并行计算, 算法的并行性得到增强.数值仿真与实际数据的实验结果表明, 本文算法与原半覆盖反投影滤波算法的重建精度相当, 但计算效率提高了4.6倍. 关键词： 锥束CT 半覆盖成像 反投影滤波 PI线     

Modification to first-order Born approximation for improved prediction of scattered sound from weakly scattering objects

When deriving the Fourier diffraction theorem based on the first-order Born approximation,the difference between wave number of the scattering object and that of the surrounding medium is ignored,causing substantial errors in sound scattering prediction.This paper modifies the Born approximation by taking into account the amplitude and phase changes between the scattering object and the water due to the wave number difference.By changing the radius and center position of the sampling circle in the Fourier domain,accuracy of the predicted sound scattering is improved.With the modified Born approximation,the computed far-field directional pattern of the scattered sound from a circular cylinder is in good agreement with the rigorous solution.Numerical calculations for several objects with different shapes are used to show applicability and effectiveness of the proposed method.     

基于可见光的多波段偏振图像融合新算法

采用了一种新的基于小波变换的偏振图像融合算法.首先,将两个波段中的每一波段三幅偏振图像利用小波变换分解成低频和高频部分,低频的小波系数平均值作为融合后的低频系数,高频细节系数根据不同区域特征选择方法以及对应输入图像小波系数的窗口区域方差来确定融合后高频小波系数,得到一个波段一幅图像.接着,将得到的图像再进行小波分解,采用低频图像的小波系数最小值作为融合后的低频系数,高频图像根据纹理一致性测度的纹理检测确定融合规则,用来调整高频小波系数,将来自不同图像的特征与细节融合在一起,并对融合图像质量进行了对比评价.实验结果表明,融合后的偏振图像不仅反映了场景的偏振信息,而且还包含了丰富的光谱信息,目标与背景的衬比度也得到了增强,为进一步的目标检测和识别提供了便利.     

基于星图识别的空间目标检测算法研究

从恒星背景中检测出空间目标是空间目标天基光学测量相机所要解决的关键技术难题之一。提出了一种基于星图识别的空间目标检测算法,利用背景星图的平移、旋转、比例伸缩不变性,应用星图识别原理,根据识别恒星在不同帧图像中的坐标变化估计背景运动参数,进行探测图像序列的配准,然后在配准后的图像序列中检测出空间目标。通过仿真验证,该方法对于运动恒星背景中运动小目标的检测具有比较好的效果。     

Efficient compression of motion-compensated sub-images with Karhunen-Loeve transform in three-dimensional integral imaging

An approach to highly enhance the compression efficiency of the integral images by applying the Karhunen-Loeve transform (KLT) algorithm to the motion-compensated sub-images is proposed. The sub-images transformed from the elemental images picked-up from the three-dimensional (3D) object might represent the different perspectives of the object. Thus, the similarity among the sub-images gets better than that among the elemental images, so that an improvement of compression efficiency of the sub-images could be obtained. However, motion vectors occurred among the sub-images might result in an additional increase of image data to be compressed. Accordingly, in this paper, motion vectors have been estimated and compensated in all sub-image in advance. Then the KLT algorithm was applied to these motion-compensated sub-images for compression. It is shown from some experimental results that compression efficiency of the proposed method has been improved up to 24.44%, 40.62%, respectively, on the average compared to that of the conventional KLT compression method and that of the JPEG.     

改进的大锥角锥束CT圆轨迹反投影滤波算法北大核心CSCD

锥束CT圆轨迹反投影滤波(BPF)算法能够有效处理数据截断问题,在感兴趣区域成像研究中具有重要的地位。但是当锥角增大时,重建图像的顶端和底端会出现严重的伪影,从而影响图像质量。为解决该问题,从扇束圆轨迹BPF算法出发,重新推导了锥束圆轨迹BPF算法,通过对锥束伪影成因分析,提出了一种基于锥角大小和轴向距离大小的权重函数,修正了重建过程中差分反投影图像求解部分,达到抑制伪影的目的。实验结果表明,改进的大锥角锥束CT圆轨迹BPF算法能够在锥角增大时有效地抑制锥角效应,提高重建图像质量。     

X光图像中缺陷的自动提取方法研究

针对炭素制品X光图像的特点,对其缺陷的提取技术进行了研究,提出了基于迭代的阈值构造方法和基于数学形态学的边缘提取算法。为快速准确地提取缺陷,设计了目标边界提取算法和基于小波变换的图像增强算法,实现了原始图像中目标区域的增强及其背景的去除。在此基础上,为排除噪声干扰的影响,采用数学形态学和迭代阈值分割相结合的方法从目标区域中提取出缺陷区域,并在迭代阈值分割的基础上,利用基于数学形态学的边缘提取算法提取了缺陷的边缘。实验结果表明,该法很好地实现了缺陷区域及其边缘的自动提取,且受噪声影响很小,为进一步的缺陷特征参量的提取与选择奠定了良好的基础。     

基于序列图像的运动目标检测

针对动态环境下运动目标检测中噪声多、目标检测不完整等情况,提出了一种基于金字塔多分辨率模型的运动目标检测方法,在低分辨率下获取目标的区域,在高分辨率下获取目标的细节。对于复杂的环境,还提出了运用高低双阈值替代传统的单阈值进行图像差分运算的方案,阈值可以根据图像自动分析得到。该方法首先将当前帧和背景帧进行尺度变换,得到不同分辨率下的图像组,然后在不同尺度下得到高低阈值差分图像,最后从高层向低层进行有效融合,得到噪声少的完整目标图像。实验表明,该方法提取运动目标的精度比较高,单目标达到0.802,多目标达到0.615,尤其是在复杂的动态环境下,优势比较明显。