Existing frequency-domain-oriented methods of parameter identification for uniform linear motion blur （ULMB） images usually dealt with special scenarios. For example, blur-kernel directions were horizontal or vertic...Existing frequency-domain-oriented methods of parameter identification for uniform linear motion blur （ULMB） images usually dealt with special scenarios. For example, blur-kernel directions were horizontal or vertical, or degraded images were of foursquare dimension. This excludes those identification methods from being applied to real images, especially to estimate undersized or oversized blur kernels. Pointing against the limitations of blur-kernel identifications, discrete Fourier transform （DFT）-based blur-kernel estimation methods are proposed in this paper. We analyze in depth the Fourier frequency response of generalized ULMB kernels, demonstrate in detail its related phase form and properties thereof, and put forward the concept of quasi-cepstrum. On this basis, methods of estimating ULMB-kernel parameters using amplitude spectrum and quasi-cepstrum are presented, respectively. The quasi-cepstrum-oriented approach increases the identifiable blur-kernel length, up to a maximum of half the diagonal length of the image. Meanwhile, directing toward the image of undersized ULMB, an improved method based on quasi-cepstrum is presented, which ameliorates the identification quality of undersized ULMB kernels. The quasi-cepstrum-oriented approach popularizes and applies the simulation-experiment- focused DFT theory to the estimation of real ULMB images. Compared against the amplitude-spectrum-oriented method, the quasi-cepstrum-oriented approach is more convenient and robust, with lower identification errors and of better noiseimmunity.展开更多
Different from a general density estimation,the crime density estimation usually has one important factor:the geographical constraint.In this paper,a new crime density estimation model is formulated,in which the regio...Different from a general density estimation,the crime density estimation usually has one important factor:the geographical constraint.In this paper,a new crime density estimation model is formulated,in which the regions where crime is impossible to happen,such as mountains and lakes,are excluded.To further optimize the estimation method,a learning-based algorithm,named Plug-and-Play,is implanted into the augmented Lagrangian scheme,which involves an off-the-shelf filtering operator.Different selections of the filtering operator make the algorithm correspond to several classical estimation models.Therefore,the proposed Plug-and-Play optimization based estimation algorithm can be regarded as the extended version and general form of several classical methods.In the experiment part,synthetic examples with different invalid regions and samples of various distributions are first tested.Then under complex geographic constraints,we apply the proposed method with a real crime dataset to recover the density estimation.The state-of-the-art results show the feasibility of the proposed model.展开更多
This paper discuss band-limited scaling function, especially on the interval band case and three interval bands case, its relationship to oversampling property and weakly translation invariance are also studied. At th...This paper discuss band-limited scaling function, especially on the interval band case and three interval bands case, its relationship to oversampling property and weakly translation invariance are also studied. At the end, we propose an open problem.展开更多
基金supported in part by the National Natural Science Foundation of China under Grant Nos. 61032007, 60972126 and 60921061the Joint Funds of the National Natural Science Foundation of China under Grant No. U0935002/L05the Natural Science Foundation of Beijing under Grant No. 4102060
文摘Existing frequency-domain-oriented methods of parameter identification for uniform linear motion blur （ULMB） images usually dealt with special scenarios. For example, blur-kernel directions were horizontal or vertical, or degraded images were of foursquare dimension. This excludes those identification methods from being applied to real images, especially to estimate undersized or oversized blur kernels. Pointing against the limitations of blur-kernel identifications, discrete Fourier transform （DFT）-based blur-kernel estimation methods are proposed in this paper. We analyze in depth the Fourier frequency response of generalized ULMB kernels, demonstrate in detail its related phase form and properties thereof, and put forward the concept of quasi-cepstrum. On this basis, methods of estimating ULMB-kernel parameters using amplitude spectrum and quasi-cepstrum are presented, respectively. The quasi-cepstrum-oriented approach increases the identifiable blur-kernel length, up to a maximum of half the diagonal length of the image. Meanwhile, directing toward the image of undersized ULMB, an improved method based on quasi-cepstrum is presented, which ameliorates the identification quality of undersized ULMB kernels. The quasi-cepstrum-oriented approach popularizes and applies the simulation-experiment- focused DFT theory to the estimation of real ULMB images. Compared against the amplitude-spectrum-oriented method, the quasi-cepstrum-oriented approach is more convenient and robust, with lower identification errors and of better noiseimmunity.
基金the National Natural Science Foundation of China under Grant Nos.61772389 and 61871260the Open Project of National Engineering Laboratory for Forensic Science of China under Grant No.2017NELKFKT02the Key Scientific Research Projects in Henan Colleges and Universities of China under Grant No.19A110015.
文摘Different from a general density estimation,the crime density estimation usually has one important factor:the geographical constraint.In this paper,a new crime density estimation model is formulated,in which the regions where crime is impossible to happen,such as mountains and lakes,are excluded.To further optimize the estimation method,a learning-based algorithm,named Plug-and-Play,is implanted into the augmented Lagrangian scheme,which involves an off-the-shelf filtering operator.Different selections of the filtering operator make the algorithm correspond to several classical estimation models.Therefore,the proposed Plug-and-Play optimization based estimation algorithm can be regarded as the extended version and general form of several classical methods.In the experiment part,synthetic examples with different invalid regions and samples of various distributions are first tested.Then under complex geographic constraints,we apply the proposed method with a real crime dataset to recover the density estimation.The state-of-the-art results show the feasibility of the proposed model.
文摘This paper discuss band-limited scaling function, especially on the interval band case and three interval bands case, its relationship to oversampling property and weakly translation invariance are also studied. At the end, we propose an open problem.
