共查询到20条相似文献,搜索用时 125 毫秒
1.
提出一种将小波仿射不变函数与凸壳相结合进行目标识别的方法。基于小波仿射不变函数构造了绝对小波仿射不变函数,作为待识别目标的特征向量,与凸壳相结合进行目标识别。该方法首先提取待识别目标的轮廓线,然后以轮廓线的一个凸壳顶点为起始点重新构造目标轮廓线点列,计算轮廓线点列的绝对小波仿射不变函数,与模板库中的模板进行匹配,根据最大相关原则识别目标。实验结果表明了该方法的有效性。 相似文献
2.
3.
4.
图像序列中运动目标的遮挡检测和处理是目前在动态图像处理时经常碰到,且比较难于解决的问题。本文针对复杂背景下图像序列的特性,给出了一种基于仿射不变量的多运动目标的遮挡检测和跟踪方法。该法首先给出了仿射不变量向量的构造,以及基于它的遮挡判决准则;然后,结合被跟踪目标的先验知识,给出丁,轨迹线以及遮挡区域的预测方法;最后,给出了依赖于仿射不变量向量的运动目标遮挡和跟踪的详细算法。实验表明新算法较好的解决了复杂背景下序列图像中的运动目标遮挡检测和跟踪的问题,较大的提高了运动目标跟踪算法的鲁棒性;同时新算法具有较快的响应速度。 相似文献
5.
本文提出了一种基于独立元分析(Independent Component Analysis,ICA)的仿射目标识别和仿射参数估计的新方法.对于待识别的不同扭曲程度的目标,可采用ICA方法提取轮廓的仿射不变描述,从而解决扭曲目标识别问题.该方法同时能够估计两个视频帧间的仿射变换参数,在高压缩率编码标准中,如MPEG4或MPEG7,具有很重要的现实意义.实验结果表明,该方法优于传统的目标识别和仿射参数估计方法. 相似文献
6.
现有的基于小波变换的形状识别算法具有很高的计算复杂度,难以满足许多实时应用的要求。文中提出了基于级联仿射不变函数的快速形状识别算法,用于识别仿射变换下的含噪目标。利用目标轮廓的小波变换可以得到一组仿射不变函数,并进一步构造出级联仿射不变函数。为了保证级联仿射不变函数的平移不变性,预先对轮廓的起始点进行了有效配准。从而通过级联仿射不变函数的内积,方便地度量出目标形状的相似度。与现有基于小波的识别算法相比,所提出的算法具有很低的计算复杂度,其所需CPU时间仅为其它算法的1/7。实验结果验证了该算法的有效性和起始点配准的准确性。 相似文献
7.
8.
针对基于传统梯度方向直方图特征的目标识别算法(HOG+SVM)在目标发生仿射变化时识别效果较差的问题,该文提出一种基于仿射梯度方向直方图特征的目标识别算法(AHOG+SVM).通过提取多尺度金字塔梯度图像的 HOG 特征,提高了算法的尺度不变性;通过将平面 HOG 栅格拓展至3维 HOG 栅格,并根据目标的世界坐标系与图像坐标系的映射关系将3维 HOG 栅格映射为2维 HOG 仿射栅格,最后对仿射栅格内的 HOG 特征进行仿射逆变换,以达到增强算法旋转不变性与错切不变性的目的.多组实验结果表明,该文提出的算法能够解决在目标识别过程中由尺度变化、旋转变化和错切变化(3D 视角变化)所造成的识别率较低的问题,性能优于 HOG+SVM算法. 相似文献
9.
10.
三维入脸识别过程及系统功能 国外三维人脸识别的典型方法主要是利用深度图像自身的几何特征,利用深度图像处理技术,分析面貌曲面的曲率等几何特征,对面貌曲面进行凹凸区域的分割、正侧面轮廓边缘的提取.最早对三维图像面貌识别的研究有Lapreste提出的基于轮廓线的方法,通过对人脸面貌曲率的分析,提取轮廓线上的特征点,利用轮廓线作为特征进行面貌的识别. 相似文献
11.
This paper is concerned with the problem of feature point registration and scene recognition from images under weak perspective transformations which are well approximated by affine transformations and under possible occlusion and/or appearance of new objects. It presents a set of local absolute affine invariants derived from the convex hull of scattered feature points (e.g., fiducial or marking points, corner points, inflection points, etc.) extracted from the image. The affine invariants are constructed from the areas of the triangles formed by connecting three vertices among a set of four consecutive vertices (quadruplets) of the convex hull, and hence do make direct use of the area invariance property associated with the affine transformation. Because they are locally constructed, they are very well suited to handle the occlusion and/or appearance of new objects. These invariants are used to establish the correspondences between the convex hull vertices of a test image with a reference image in order to undo the affine transformation between them. A point matching approach for recognition follows this. The time complexity for registering L feature points on the test image with N feature points of the reference image is of order O(NxL). The method has been tested on real indoor and outdoor images and performs well. 相似文献
12.
Xiang-yang Wang Pan-pan Niu Hong-ying Yang Li-li Chen 《Journal of Visual Communication and Image Representation》2012,23(6):892-907
Feature point based image watermarking against geometric distortions has attracted great attention in recent years. However, for the state-of-the-art intensity based feature points detectors, the feature points often gather at textured portions of the image or on the edges where the change of intensity is significant, so that many feature points capture the same portion of the image, which makes the watermark be vulnerable to local geometric distortions. In this paper, we propose an affine invariant image watermarking scheme with good visual quality and reasonable resistance toward local geometric distortions, which utilizes the intensity probability density-based Harris–Laplace detector. Firstly, the uniform and robust feature points are extracted by utilizing modified Harris–Laplace detector, in which the intensity probability density gradient is used instead of intensity gradient. Then, the affine invariant local ellipse regions (LERs) are constructed adaptively according to the variation of local intensity probability density. Finally, the digital watermark is embedded into the affine invariant LERs in nonsubsampled contourlet transform (NSCT) domain by modulating the lowpass NSCT coefficients. By binding the watermark with the affine invariant LERs, the watermark detection can be done without synchronization error. Experimental results show that the proposed image watermarking is not only invisible and robust against common image processing operations such as sharpening, noise adding, and JPEG compression, but also robust against the global affine transforms and local geometric distortions. 相似文献
13.
A new approach is proposed for registering a set of histological coronal two-dimensional images of a rat brain sectional material with coronal sections of a three-dimensional brain atlas, an intrinsic step and a significant challenge to current efforts in brain mapping and multimodal fusion of experimental data. The alignment problem is based on matching external contours of the brain sections, and operates in the presence of tissue distortion and tears which are routinely encountered, and possible scale, rotation, and shear changes (the affine and weak perspective groups). It is based on a novel set of local absolute affine invariants derived from the set of ordered inflection points on the external contour represented by a cubic B-spline curve. The inflection points are local intrinsic geometric features, which are preserved under both the affine and the weak perspective transformations. The invariants are constructed from the sequence of area patches bounded by the contour and the line connecting two consecutive inflection points, and hence do make direct use of the area (volume) invariance property associated with the affine transformation. These local absolute invariants are very well suited to handle the tissue distortion and tears (occlusion problem). 相似文献
14.
为了解决SAR图像受相干斑噪声干扰和震后发生形变而识别率偏低的问题,提出了一种新的仿射、形变不变特征-热核特征,并将该特征用于SAR图像目标识别.首先采用推广的核模糊C-均值方法分割SAR图像,提取SAR图像目标形状;接着对目标形状进行Delaunay三角剖分,采用余切权重法对Laplace-Beltrami Operator离散化,通过离散化Laplace-Beltrami Operator特征值、特征向量求每一点热核特征;然后采用谱距离公式对点点间热核距离计算,转化为距离分布表示目标形状的热核特征;最后采用L1相似性准则对图像进行相似性度量,得到识别结果.实验表明:与经典的Hu不变矩方法相比,对于仿射变换和发生形变的SAR图像,该方法都具有更高的识别率.因此,基于热核特征的SAR图像识别方法是一种更加有效的识别方法. 相似文献
15.
针对传统基于梯度方向直方图特征检测算法对解决目标模型单一、发生形变、存在遮挡及目标受干扰下定位困难的问题,提出一种基于HOG特征混合模型结合隐SVM的感兴趣目标检测算法。首先利用用训练图像的HOG特征金字塔表示得到包含感兴趣目标根模型、部件模型和对应可变形部件特征表示,该模型不仅描述目标的整体轮廓,而且能够捕捉到更为精细的目标部件轮廓,在一定程度上提高了检测算法在目标姿态复杂情况下的鲁棒性。然后利用HOG特征混合特征训练部件检测分类器LSVM(Latent Support Vector Machine)。最后通过动态规划和距离转换算法在测试图上扫描出与可变形部件模型相匹配的区域,实现感兴趣目标的检测定位。经过多组实验结果表明,所提出的算法能较好地解决目标在发生较大形变和存在遮挡等复杂姿态下的定位问题。 相似文献
16.
17.
A sub-pixel disparity refinement algorithm based on Lagrange interpolation is proposed to calculate the sub-pixel disparity and the elevation of the target con-struction on the remote sensing image. The accurate in-teger matching feature points are obtained by improved affine scale invariant feature transform. A series of inter-polation samples can be estimated through the three times magnification results of each feature matching point and its close neighbor pixels. The final sub-pixel deviation can be calculated by the interpolation samples. Experiment re-sults show that a more accurate elevation information can be obtained on the remote sensing image with corrective epipolar line, and an accurate sub-pixel disparity result can be estimated without the epipolar line constraint as well. 相似文献
18.
该文给出了一种基于稀疏特征匹配和形变传播的无缝图像拼接方法。首先,在配准图像的重叠区域中寻找一条结构误差最小的最佳接缝,从一边的图像中选取目标区域;接着,沿着接缝在两边的图像区域中检测出显著的结构特征,并进行特征匹配,获得目标区域中接缝上的匹配特征点及与之关联的边缘特征点的结构形变矢量;然后,通过求解泊松方程,将这些稀疏的形变矢量稳定和平滑地传播到目标区域内部,得到目标区域中各点的形变矢量;最后,由形变矢量通过内插获得目标区域的梯度场,并由梯度场重构出最终结果。该方法执行方便快速,不需要复杂的特征检测,能够统一地纠正图像拼接中较大的结构错位和颜色过渡不自然,在全局上消除结构接缝和颜色接缝。与其它方法比较,该方法获得较明显的改进。 相似文献
19.
针对视频序列中目标的跟踪问题,提出了一种基于粒子滤波框架的联合仿射和外貌模型的目标跟踪算法.该算法首先提取图像帧之间的相关特征点,通过求解Sylvester方程得到仿射参数,然后将仿射参数嵌入到基于仿射群的粒子滤波框架中进行平滑估计.利用基于仿射群的一阶自回归过程模拟状态的变化,联合仿射特征点模型和外貌模型进行似然估计,得到粒子的最佳平均状态,进而对目标实施跟踪.实验结果表明,在目标经历姿势和尺度变化、遮挡以及复杂背景等情况下,提出的算法能够有效地跟踪目标,较之其他相关算法具有很强的鲁棒性. 相似文献