Feature-based image matching algorithms play an indispensable role in automatic target recognition （ATR）. In this work, a fast image matching algorithm （FIMA） is proposed which utilizes the geometry feature of ext...Feature-based image matching algorithms play an indispensable role in automatic target recognition （ATR）. In this work, a fast image matching algorithm （FIMA） is proposed which utilizes the geometry feature of extended centroid （EC） to build affine invariants. Based on at-fine invariants of the length ratio of two parallel line segments, FIMA overcomes the invalidation problem of the state-of-the-art algorithms based on affine geometry features, and increases the feature diversity of different targets, thus reducing misjudgment rate during recognizing targets. However, it is found that FIMA suffers from the parallelogram contour problem and the coincidence invalidation. An advanced FIMA is designed to cope with these problems. Experiments prove that the proposed algorithms have better robustness for Gaussian noise, gray-scale change, contrast change, illumination and small three-dimensional rotation. Compared with the latest fast image matching algorithms based on geometry features, FIMA reaches the speedup of approximate 1.75 times. Thus, FIMA would be more suitable for actual ATR applications.展开更多
Expression, occlusion, and pose variations are three main challenges for 3D face recognition. A novel method is presented to address 3D face recognition using scale-invariant feature transform(SIFT) features on 3D mes...Expression, occlusion, and pose variations are three main challenges for 3D face recognition. A novel method is presented to address 3D face recognition using scale-invariant feature transform(SIFT) features on 3D meshes. After preprocessing, shape index extrema on the 3D facial surface are selected as keypoints in the difference scale space and the unstable keypoints are removed after two screening steps. Then, a local coordinate system for each keypoint is established by principal component analysis(PCA).Next, two local geometric features are extracted around each keypoint through the local coordinate system. Additionally, the features are augmented by the symmetrization according to the approximate left-right symmetry in human face. The proposed method is evaluated on the Bosphorus, BU-3DFE, and Gavab databases, respectively. Good results are achieved on these three datasets. As a result, the proposed method proves robust to facial expression variations, partial external occlusions and large pose changes.展开更多
In this report several practical issues about moment invariants with application to image classification are concerned. A modified formulation for the approximation of the moments of digital images is suggested. Four ...In this report several practical issues about moment invariants with application to image classification are concerned. A modified formulation for the approximation of the moments of digital images is suggested. Four computational procedures and their corresponding noise performances are studied in detail.展开更多
For vision-based mobile robot navigation, images of the same scene may undergo a general affine transformation in the case of significant viewpoint changes. So, a novel method for detecting affine invariant interest p...For vision-based mobile robot navigation, images of the same scene may undergo a general affine transformation in the case of significant viewpoint changes. So, a novel method for detecting affine invariant interest points is proposed to obtain the invariant local features, which is coined polynomial local orientation tensor(PLOT). The new detector is based on image local orientation tensor that is constructed from the polynomial expansion of image signal. Firstly, the properties of local orientation tensor of PLOT are analyzed, and a suitable tuning parameter of local orientation tensor is chosen so as to extract invariant features. The initial interest points are detected by local maxima search for the smaller eigenvalues of the orientation tensor. Then, an iterative procedure is used to allow the initial interest points to converge to affine invariant interest points and regions. The performances of this detector are evaluated on the repeatability criteria and recall versus 1-precision graphs, and then are compared with other existing approaches. Experimental results for PLOT show strong performance under affine transformation in the real-world conditions.展开更多
Synthetic aperture radar(SAR)is a high-resolution two-dimensional imaging radar.However,during the imaging process,SAR is susceptible to intentional and unintentional interference,with radio frequency inter⁃ference(RF...Synthetic aperture radar(SAR)is a high-resolution two-dimensional imaging radar.However,during the imaging process,SAR is susceptible to intentional and unintentional interference,with radio frequency inter⁃ference(RFI)being the most common type,leading to a severe degradation in image quality.To address the above problem,numerous algorithms have been proposed.Although inpainting networks have achieved excellent results,their generalization is unclear.Whether they still work effectively in cross-sensor experiments needs fur⁃ther verification.Through the time-frequency analysis to interference signals,this work finds that interference holds domain invariant features between different sensors.Therefore,this work reconstructs the loss function and extracts the domain invariant features to improve its generalization.Ultimately,this work proposes a SAR RFI suppression method based on domain invariant features,and embeds the RFI suppression into SAR imaging pro⁃cess.Compared to traditional notch filtering methods,the proposed approach not only removes interference but also effectively preserves strong scattering targets.Compared to PISNet,our method can extract domain invariant features and hold better generalization ability,and even in the cross-sensor experiments,our method can still achieve excellent results.In cross-sensor experiments,training data and testing data come from different radar platforms with different parameters,so cross-sensor experiments can provide evidence for the generalization.展开更多
跨域图像识别技术旨在利用已标注源域完成未标注目标域上的图像标注任务,且这2个域服从不同数据分布。现有方法一般通过提取域不变表征缩小域间分布差异,从而提升源域训练模型在目标域上的图像标注能力。然而,部分源域样本与目标域样本...跨域图像识别技术旨在利用已标注源域完成未标注目标域上的图像标注任务,且这2个域服从不同数据分布。现有方法一般通过提取域不变表征缩小域间分布差异,从而提升源域训练模型在目标域上的图像标注能力。然而,部分源域样本与目标域样本的分布偏差较大,强行提取所有源域样本的域不变表征容易导致模型负适配问题。为此,提出一种基于类条件最优传输的跨域样本选择(CDSS)策略。对于2个域中对应相同类的样本,该策略在源域中选择最能代表目标域的样本子集,并将该过程建模为在源域上学习一个稀疏概率分布的优化问题,使该分布与目标域分布之间的最优传输距离最小。在3个跨域图像识别数据集上的实验结果表明,应用所提策略后,方法的平均识别精度较3种主流基准方法(批谱正则化项(BSP)、批量核范数最大化(BNM)和MDD(Maximum Density Divergence))均取得了显著提升,验证了所提策略在改善模型负适配问题上的有效性。展开更多
Image matching based on scale invariant feature transform（SIFT） is one of the most popular image matching algorithms, which exhibits high robustness and accuracy. Grayscale images rather than color images are genera...Image matching based on scale invariant feature transform（SIFT） is one of the most popular image matching algorithms, which exhibits high robustness and accuracy. Grayscale images rather than color images are generally used to get SIFT descriptors in order to reduce the complexity. The regions which have a similar grayscale level but different hues tend to produce wrong matching results in this case. Therefore, the loss of color information may result in decreasing of matching ratio. An image matching algorithm based on SIFT is proposed, which adds a color offset and an exposure offset when converting color images to grayscale images in order to enhance the matching ratio. Experimental results show that the proposed algorithm can effectively differentiate the regions with different colors but the similar grayscale level, and increase the matching ratio of image matching based on SIFT. Furthermore, it does not introduce much complexity than the traditional SIFT.展开更多
基金Projects(2012AA010901,2012AA01A301)supported by National High Technology Research and Development Program of ChinaProjects(61272142,61103082,61003075,61170261,61103193)supported by the National Natural Science Foundation of ChinaProjects(B120601,CX2012A002)supported by Fund Sponsor Project of Excellent Postgraduate Student of NUDT,China
文摘Feature-based image matching algorithms play an indispensable role in automatic target recognition （ATR）. In this work, a fast image matching algorithm （FIMA） is proposed which utilizes the geometry feature of extended centroid （EC） to build affine invariants. Based on at-fine invariants of the length ratio of two parallel line segments, FIMA overcomes the invalidation problem of the state-of-the-art algorithms based on affine geometry features, and increases the feature diversity of different targets, thus reducing misjudgment rate during recognizing targets. However, it is found that FIMA suffers from the parallelogram contour problem and the coincidence invalidation. An advanced FIMA is designed to cope with these problems. Experiments prove that the proposed algorithms have better robustness for Gaussian noise, gray-scale change, contrast change, illumination and small three-dimensional rotation. Compared with the latest fast image matching algorithms based on geometry features, FIMA reaches the speedup of approximate 1.75 times. Thus, FIMA would be more suitable for actual ATR applications.
基金Project(XDA06020300)supported by the"Strategic Priority Research Program"of the Chinese Academy of SciencesProject(12511501700)supported by the Research on the Key Technology of Internet of Things for Urban Community Safety Based on Video Sensor networks
文摘Expression, occlusion, and pose variations are three main challenges for 3D face recognition. A novel method is presented to address 3D face recognition using scale-invariant feature transform(SIFT) features on 3D meshes. After preprocessing, shape index extrema on the 3D facial surface are selected as keypoints in the difference scale space and the unstable keypoints are removed after two screening steps. Then, a local coordinate system for each keypoint is established by principal component analysis(PCA).Next, two local geometric features are extracted around each keypoint through the local coordinate system. Additionally, the features are augmented by the symmetrization according to the approximate left-right symmetry in human face. The proposed method is evaluated on the Bosphorus, BU-3DFE, and Gavab databases, respectively. Good results are achieved on these three datasets. As a result, the proposed method proves robust to facial expression variations, partial external occlusions and large pose changes.
文摘In this report several practical issues about moment invariants with application to image classification are concerned. A modified formulation for the approximation of the moments of digital images is suggested. Four computational procedures and their corresponding noise performances are studied in detail.
基金Projects(61203332,61203208) supported by the National Natural Science Foundation of China
文摘For vision-based mobile robot navigation, images of the same scene may undergo a general affine transformation in the case of significant viewpoint changes. So, a novel method for detecting affine invariant interest points is proposed to obtain the invariant local features, which is coined polynomial local orientation tensor(PLOT). The new detector is based on image local orientation tensor that is constructed from the polynomial expansion of image signal. Firstly, the properties of local orientation tensor of PLOT are analyzed, and a suitable tuning parameter of local orientation tensor is chosen so as to extract invariant features. The initial interest points are detected by local maxima search for the smaller eigenvalues of the orientation tensor. Then, an iterative procedure is used to allow the initial interest points to converge to affine invariant interest points and regions. The performances of this detector are evaluated on the repeatability criteria and recall versus 1-precision graphs, and then are compared with other existing approaches. Experimental results for PLOT show strong performance under affine transformation in the real-world conditions.
基金Supported by the National Natural Science Foundation of China(62001489)。
文摘Synthetic aperture radar(SAR)is a high-resolution two-dimensional imaging radar.However,during the imaging process,SAR is susceptible to intentional and unintentional interference,with radio frequency inter⁃ference(RFI)being the most common type,leading to a severe degradation in image quality.To address the above problem,numerous algorithms have been proposed.Although inpainting networks have achieved excellent results,their generalization is unclear.Whether they still work effectively in cross-sensor experiments needs fur⁃ther verification.Through the time-frequency analysis to interference signals,this work finds that interference holds domain invariant features between different sensors.Therefore,this work reconstructs the loss function and extracts the domain invariant features to improve its generalization.Ultimately,this work proposes a SAR RFI suppression method based on domain invariant features,and embeds the RFI suppression into SAR imaging pro⁃cess.Compared to traditional notch filtering methods,the proposed approach not only removes interference but also effectively preserves strong scattering targets.Compared to PISNet,our method can extract domain invariant features and hold better generalization ability,and even in the cross-sensor experiments,our method can still achieve excellent results.In cross-sensor experiments,training data and testing data come from different radar platforms with different parameters,so cross-sensor experiments can provide evidence for the generalization.
文摘跨域图像识别技术旨在利用已标注源域完成未标注目标域上的图像标注任务,且这2个域服从不同数据分布。现有方法一般通过提取域不变表征缩小域间分布差异,从而提升源域训练模型在目标域上的图像标注能力。然而,部分源域样本与目标域样本的分布偏差较大,强行提取所有源域样本的域不变表征容易导致模型负适配问题。为此,提出一种基于类条件最优传输的跨域样本选择(CDSS)策略。对于2个域中对应相同类的样本,该策略在源域中选择最能代表目标域的样本子集,并将该过程建模为在源域上学习一个稀疏概率分布的优化问题,使该分布与目标域分布之间的最优传输距离最小。在3个跨域图像识别数据集上的实验结果表明,应用所提策略后,方法的平均识别精度较3种主流基准方法(批谱正则化项(BSP)、批量核范数最大化(BNM)和MDD(Maximum Density Divergence))均取得了显著提升,验证了所提策略在改善模型负适配问题上的有效性。
基金supported by the National Natural Science Foundation of China(61271315)the State Scholarship Fund of China
文摘Image matching based on scale invariant feature transform（SIFT） is one of the most popular image matching algorithms, which exhibits high robustness and accuracy. Grayscale images rather than color images are generally used to get SIFT descriptors in order to reduce the complexity. The regions which have a similar grayscale level but different hues tend to produce wrong matching results in this case. Therefore, the loss of color information may result in decreasing of matching ratio. An image matching algorithm based on SIFT is proposed, which adds a color offset and an exposure offset when converting color images to grayscale images in order to enhance the matching ratio. Experimental results show that the proposed algorithm can effectively differentiate the regions with different colors but the similar grayscale level, and increase the matching ratio of image matching based on SIFT. Furthermore, it does not introduce much complexity than the traditional SIFT.
