基于背景抑制和特征点检测的目标检测算法

空域远距离红外目标探测系统中,飞行目标多表现为点状或面状的小目标,像素数少,且常伴有低空地面物体的干扰.根据空域和地面在梯度变化上的不同和目标本身的特性,提出了一种基于地面背景抑制和特征点检测的红外空中目标检测算法.分析了地面和空域在梯度变化上的特点,根据梯度变化大的像素的整体统计信息划分了空域和地面在图像中的分布,再通过特征点检测实现了候选红外飞行目标的检测.该算法适用于纯空域和低空背景,经过对实际采集的大量红外图像的仿真表明,本文提出的算法具有很强的实用性和鲁棒性.     

基于高斯混合背景模型的运动目标检测技术研究

在运动目标检测技术中,使用传统的高斯混合背景模型所得到的检测结果并不能完美地获取运动目标的轮廓信息,而图像中像素的梯度信息,刚好就是反映了各物体的轮廓和边界,并且相对于颜色信息而言,梯度信息对于噪声并不敏感。为此,该文对传统的高斯混合背景模型进行了改进,提出基于梯度时空信息的高斯混合背景模型,证明了改进的算法确实能够取...     

基于边缘多通道梯度模型的多运动目标检测

针对视频序列运动目标检测易受环境噪声干扰、提取目标轮廓困难的问题,提出了一种基于边缘多通道梯度改进模型的多运动目标检测算法。首先,利用Canny算子获取视频序列中目标的边缘信息,并根据人类视觉色彩的恒常特性,对目标边缘建立时间、空间、颜色多通道梯度模型;然后,利用该模型获取目标边缘像素点的运动状态描述信息,实现背景边缘和运动物体边缘的分离;最后,将间断边缘像素点与其邻域点的运动状态相关联,以连接目标间断边缘,实现运动目标轮廓的提取,并将连接后的轮廓进行形态学处理以分割出目标。实验结果表明,与同类型算法相比,本算法在运动目标检测中具有的实时性、准确性和鲁棒性更好。     

基于深度学习的红外与可见光决策级融合检测

提出了一种基于深度学习的红外与可见光决策级融合检测方法。首先,提出了一种介于深度学习模型之间的参数传递模型,进而从基于深度学习的可见光物体检测模型上抽取了用于红外物体检测的预训练模型,并在课题组实地采集的红外数据集上进行fine-tuning,从而得到基于深度学习的红外物体检测模型。在此基础上,提出了一种基于深度学习的红外与可见光决策级融合检测模型,并对模型设计、图像配准、决策级融合过程进行了详细地阐述。最后,进行了白天和傍晚条件下基于深度学习的单波段检测实验和双波段融合检测实验。定性分析上,由于波段之间的信息互补性,相比于单波段物体检测,双波段融合物体检测在检测结果上具有更高的置信度和更精确的物体框;定量分析上,白天时,双波段融合检测的mAP为86.0%,相比于红外检测和可见光检测分别提高了9.9%和5.3%;傍晚时,双波段融合检测的mAP为89.4%,相比于红外检测和可见光检测分别提高了3.1%和14.4%。实验结果表明:基于深度学习的双波段融合检测方法相比于单波段检测方法具有更好的检测性能和更强的鲁棒性,同时也验证了所提出方法的有效性。     

复杂背景中圆弧检测新方法

大量物体具有圆弧轮廓,可以通过探测圆弧来识别物体的存在,以及应用于产品质量评估、在线检测和机器人视觉等领域.文章研究了基于图像传感器的复杂背景下的圆弧检测方法,提出一种圆弧检测新方法.传统的霍夫变换最大特点是可靠性高,在噪声、变形、甚至部分区域丢失的状态下仍然能够取得理想的结果,但缺点是计算复杂度高,资源需求大.本文所提出的方法降低了计算复杂性和对资源的需求.实验结果表明该方法快速、准确、有效.     

大曲率表面物体形貌的光学检测方法研究

本文提出了两种用于曲面形貌检测的傅里叶变换法中的频域低通滤波器，并与常用的窗口法低通滤波器进行了比较，给出了实际测量的结果。这两种频域低通滤波器（高幂次指数滤波器和矩形余弦滤波器）有可方便调整且平直的通带，平直无旁瓣的阻带，较陡的边带将通带与阻带间圆滑过渡连接，从而可尽可能多地保留有用频率成分，很好地滤掉直流和不用的高频分量，且较好地抑制振铃现象，克服一般窗口法设计的低通滤波器中旁瓣的影响。具有使用简单、参数调整方便、应用范围广等特点，尤其是对具有大梯度表面和棱角细节的物体检测更有独到之处，具有广泛的应用前景。     

基于监控视频中背景重组的运动物体的检测

智能监控视频中,核心技术是运动物体视觉分析。运动物体的检测又是运动物体视觉分析的基础和关键。对于运动物体的检测与跟踪,背景的提取干净与否关系到后续工作的准确性。文中提出了新的背景确立方法,以及在运动物体的检测方面也提出了新的改进方法。实验表明该方法具有一定的有效性,实用性。     

一种基于梯度的细胞图像边缘检测算法

医学图像处理中对细胞形态的分析主要依赖于边缘检测技术,针对医学细胞图像的特性,提出了一种新的边缘检测算法,算法基于梯度原理,采用5×5检测窗口,分别计算水平、垂直方向的4个条形子窗口的像素均值差,采用它们绝对值的最大值做为各自的方向梯度,再取两个方向梯度的最大者作为检测窗口中心点的梯度值,最后对梯度图像进行细化并提取边缘。实验证明,提出的算法提取的细胞边缘连续性好、精确度高,具有一定的抗噪声能力,性能优于传统的检测算法。     

基于邻近轮廓片段的物体检测

本文提出一种新颖的物体特征：邻近轮廓片段。该特征能够很好的描述物体边缘特征，而且具有尺度不变性。利用该特征训练得到的SVM检测器，在物体检测过程中取得了很好的效果。经实验验证，该特征非常适合用于物体检测。     

基于DBSCAN与梯度划分的Kinect障碍物轮廓检测算法

针对移动机器人的环境检测和避障问题中传感器获取的信息不够全面及准确,无法准确提供周围环境信息等问题,文中提出了利用Kinect传感器来获取周围环境的色彩信息和深度数据,并且提出了一种利用梯度划分和DBSCAN聚类方法来处理Kinect传感器获得的深度数据图。该算法首先使用梯度障碍物边缘检测方法对Kinect获取得到的深度图进行快速高效的处理障碍物边缘轮廓,并对算法中的差分参数进行改进,使得计算得到的梯度结果更准确。然后对比不同的聚类方法,使用BDSCAN聚类方法来对检测划分完毕的障碍物进行聚类分析,最后通过安排具体实验对该算法进行验证。实验结果表明,该算法能够对周围环境障碍物进行准确划分,可行区域效果明显,对不同物体的成功检测率较高,验证了算法的有效性。     

Gradient Estimation Using Wide Support Operators

One of the fastest methods of localizing edges in images is based on small gradient kernels, such as Sobel, Prewitt, and Roberts. Although small gradient kernels provide a fast way of computing the gradients, they have little control over noise, edge location, and edge orientation. They are known to be only sensitive to step edges and fail to detect smooth boundaries. On the other hand, large kernels provide superior noise suppression characteristics, but they suffer from wide response area around edges. They cause edges of neighboring objects to merge due to their wide support. Problems associated with large gradient kernels prevent their widespread usage. This paper presents a fuzzy topology-based method to facilitate the use of larger gradient kernels. The new method effectively limits the response area around the edge and prevents neighboring objects to affect each other. Synthetic images are used to show the superior noise suppression properties and response characteristics to both step and ramp edges. Natural images are also used to assess the performance of the newly proposed topological gradient estimation qualitatively.      

Simulation of near-surface detection of objects in layered media by the BCGS-FFT method

Near-surface electromagnetic characterization of objects buried in multilayered earth is important for the detection and identification of landmines, unexploded ordnance, and underground structures. However, so far little progress has been made in the development of fast algorithms for inhomogeneous objects in a layered medium. We report an iterative technique, the stabilized biconjugate gradient fast Fourier transform (BCGS-FFT) method, that simulates near-surface detection of three-dimensional, inhomogeneous objects buried in multilayered media. The CPU time and memory cost of the BCGS-FFT method is O(NlogN) and O(N), respectively, where N is the number of unknowns. This method is significantly more efficient than method of moments. It is capable of solving large-scale electromagnetic scattering problems with an arbitrary inhomogeneous object embedded in a layered medium with an arbitrary number of layers. Examples in subsurface detection of large buried objects are shown to demonstrate the efficacy of this method. At present, the object must be located completely within one single layer in this multilayer medium, but efforts are underway to remove this limitation.     

彩色图像序列中运动目标外轮廓的自动提取

该文提出了一种新的多运动目标精确外轮廓自动提取算法。算法的主要创新点是:采用指数滤波计算彩色图像的梯度强度,改善了彩色图像梯度局部极值点的定位精度,所得到的梯度强度和运动边缘位置信息供随后的轮廓迭代中使用,成功地避免了动态轮廓的过收缩。此外,充分利用轮廓的方向信息,又有效地克服了动态轮廓迭代中的外扩展。从而,算法自动地提取出一个真正紧贴的目标外轮廓。分析和实验表明,该算法抗干扰能力强,对复杂背景及不重叠物体,可以精确定位并提取出各个运动目标的外轮廓。     

Image segmentation and analysis via multiscale gradient watershedhierarchies

Multiscale image analysis has been used successfully in a number of applications to classify image features according to their relative scales. As a consequence, much has been learned about the scale-space behavior of intensity extrema, edges, intensity ridges, and grey-level blobs. We investigate the multiscale behavior of gradient watershed regions. These regions are defined in terms of the gradient properties of the gradient magnitude of the original image. Boundaries of gradient watershed regions correspond to the edges of objects in an image. Multiscale analysis of intensity minima in the gradient magnitude image provides a mechanism for imposing a scale-based hierarchy on the watersheds associated with these minima. This hierarchy can be used to label watershed boundaries according to their scale. This provides valuable insight into the multiscale properties of edges in an image without following these curves through scale-space. In addition, the gradient watershed region hierarchy can be used for automatic or interactive image segmentation. By selecting subtrees of the region hierarchy, visually sensible objects in an image can be easily constructed.     

基于面积约束和自适应梯度修正的分水岭图像分割

图像中的噪声或非规则细节干扰易导致形态学 分水岭产生较严重的过分割,为了在消除过分割的同时尽可能 保持图像目标边界的准确定位,提出了一种基于面积约束和自适应梯度修正的分水岭图像分 割方法。首先对图像进行梯 度变换,采用区域面积约束滤除狭小高梯度尖峰对应的噪声和非规则细节;然后建立梯度级 与结构元素大小之间的函数 关系,并以相对应的结构元素对梯度图像进行粘性形态学(VM)闭运算,消除低梯度噪声及非 规则细节,实现梯度图像的自适 应修正,由于VM闭运算对梯度图像进行修正时,对目标仅作轻度或不作修正,因 而能够最大限度的保持目标轮 廓的准确定位,而对噪声和非规则细节则采用较大尺寸的结构元素进行较大幅度修正,从而 消除产生过分割的因素;最 后对修正图像进行分水岭分割。实验结果表明,本文方法能够在消除过分割的同时,保持目 标轮廓的准确定位。     

Optimal orientation estimators for detection of cylindrical objects

This paper introduces low-level operators in the context of detecting cylindrical axis in 3D images. Knowing the axis of a cylinder is particularly useful since its location, length and curvature derive from this knowledge. This paper introduces a new gradient-based optimal operator dedicated to accurate estimation of the direction toward the axis. The operator relies on Finite Impulse Response filters. The approach is presented first in a 2D context, thus providing optimal gradient masks for locating the center of circular objects. Then, a 3D extension is provided, allowing the exact estimation of the orientation toward the axis of cylindrical objects when this axis coincides with one of the mask reference axes. Applied to more general cylinders and to noisy data, the operator still provides accurate estimation and outperforms classical gradient operators.     

A joint cumulative distribution function and gradient fusion based method for dehazing of long shot hazy images

Hazy or foggy weather conditions significantly degrade the visual quality of an image in an outdoor environment. It also changes the color and reduces the contrast of an image. This paper introduces a novel single image dehazing technique to restore a hazy image without considering the physical model of haze formation. In order to find haze-free image, the proposed method does not require the transmission map and its costly refinement process. Since haze effect is dependent on the depth, it severely degrades the visibility of the objects located at a far distance. The objects close to the camera are unaffected. In this paper, we propose a fusion-based haze removal method based on the joint cumulative distribution function (JCDF) that treats faraway haze and nearby haze separately. The output images after the JCDF module, fused in the gradient domain to produce a haze-free image. The proposed method not only significantly enhances visibility but also preserves texture details. The proposed method is experimented and evaluated on a large set of challenging hazy images (large scene depth, night time, dense fog, etc.). Both qualitative and quantitative measures show that the performance of the proposed method is better than the state-of-the-art dehazing techniques.     

Analysis of two-dimensional conducting and dielectric bodies utilizing the conjugate gradient method

The conjugate gradient method for the analysis of both transverse electric (TE) and transverse magnetic (TM) scattering from two-dimensional conducting and dielectric structures is presented. Comparison with the conventional approaches indicate that the conjugate gradient method yields good results. However, the major advantage of the conjugate gradient method is that one could incorporate any initial guess for the solution of the current in the solution procedure and that a given problem could be solved to any prespecified degree of accuracy. Also the conjugate gradient method may he applied to the solution of operator equations at frequencies corresponding to the internal resonances of the objects yielding a minimum norm solution. The storage required for the conjugate gradient method is5Nas opposed toN^{2}for conventional methods.     

Active contours without edges

We propose a new model for active contours to detect objects in a given image, based on techniques of curve evolution, Mumford-Shah (1989) functional for segmentation and level sets. Our model can detect objects whose boundaries are not necessarily defined by the gradient. We minimize an energy which can be seen as a particular case of the minimal partition problem. In the level set formulation, the problem becomes a "mean-curvature flow"-like evolving the active contour, which will stop on the desired boundary. However, the stopping term does not depend on the gradient of the image, as in the classical active contour models, but is instead related to a particular segmentation of the image. We give a numerical algorithm using finite differences. Finally, we present various experimental results and in particular some examples for which the classical snakes methods based on the gradient are not applicable. Also, the initial curve can be anywhere in the image, and interior contours are automatically detected.