Summary form only given. A least-squares filter for the restoration of multichannel images is presented. The process involves the removal of noise and degradation from observed multichannel imagery, such as color or multispectral images. The restoration filters utilize information distributed across image channels and process all channels as a single entity. They use a priori information and constraints, thus avoiding some of the drawbacks of the minimum-mean-squared-error filter.<>
{"title":"Digital least squares restoration of multi-channel images","authors":"N. P. Galatsanos, R. Chin","doi":"10.1109/MDSP.1989.97105","DOIUrl":"https://doi.org/10.1109/MDSP.1989.97105","url":null,"abstract":"Summary form only given. A least-squares filter for the restoration of multichannel images is presented. The process involves the removal of noise and degradation from observed multichannel imagery, such as color or multispectral images. The restoration filters utilize information distributed across image channels and process all channels as a single entity. They use a priori information and constraints, thus avoiding some of the drawbacks of the minimum-mean-squared-error filter.<<ETX>>","PeriodicalId":340681,"journal":{"name":"Sixth Multidimensional Signal Processing Workshop,","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1989-09-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130252876","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Summary form only given. An algorithm that calculates the maximum a posteriori estimate of the complete sinogram has been developed. It uses prior knowledge of the smoothness of the sinogram, fundamental mathematical constraints on the Radon transform, and a complete probabilistic characterization of the observation noise. The object is reconstructed using convolution backprojection applied to the restored sinogram. The observation that many objects of interest tend to have smooth sinograms, although the objects themselves may not be smooth, has been incorporated by defining a Markov random field prior probability on full sinograms, rather than on objects. The Markov random field used is of the simplest kind-nearest neighbor with quadratic potential terms-although more elaborate models can be used. Using a known noise model (zero-mean, Gaussian), the maximum a posteriori solution to the sinogram restoration problem can be formulated. The solution to this problem is a constrained optimization algorithm, and because of the simple form of both the prior and the observation noise, it was possible to develop an iterative primal-dual algorithm that converges quite rapidly too the desired solution.<>
{"title":"Limited-angel tomography using constrained sinogram restoration","authors":"Jerry L Prince, A. Willsky","doi":"10.1109/MDSP.1989.97125","DOIUrl":"https://doi.org/10.1109/MDSP.1989.97125","url":null,"abstract":"Summary form only given. An algorithm that calculates the maximum a posteriori estimate of the complete sinogram has been developed. It uses prior knowledge of the smoothness of the sinogram, fundamental mathematical constraints on the Radon transform, and a complete probabilistic characterization of the observation noise. The object is reconstructed using convolution backprojection applied to the restored sinogram. The observation that many objects of interest tend to have smooth sinograms, although the objects themselves may not be smooth, has been incorporated by defining a Markov random field prior probability on full sinograms, rather than on objects. The Markov random field used is of the simplest kind-nearest neighbor with quadratic potential terms-although more elaborate models can be used. Using a known noise model (zero-mean, Gaussian), the maximum a posteriori solution to the sinogram restoration problem can be formulated. The solution to this problem is a constrained optimization algorithm, and because of the simple form of both the prior and the observation noise, it was possible to develop an iterative primal-dual algorithm that converges quite rapidly too the desired solution.<<ETX>>","PeriodicalId":340681,"journal":{"name":"Sixth Multidimensional Signal Processing Workshop,","volume":"25 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1989-09-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134049558","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Summary form only given. The reconstruction of signals from phase only information has been treated. Although this is not possible in general, it has been shown previously that a finite duration sequence, provided its z-transform has no zeros in reciprocal pairs or on the unit circle, is uniquely specified by its Fourier transform phase within a scale factor and therefore can be reconstructed within a scale factor. The derivation assumes that the phases are known exactly; no error of any kind is involved. As all measured quantities are degraded by some kind of error, this assumption is much too restrictive when one wishes to reconstruct real signals as opposed to simulated signals. To study the effects of phase degradation on the quality of reconstructed signals, a series of experiments has been performed. The effects of additive random noise and of quantization noise in the phase samples have been investigated along with the effects of uniform and nonuniform sampling of the Fourier phase function. It has been found that the quality of a reconstructed signal is dependent on the choice of the set of frequencies used for sampling the Fourier phase function.<>
{"title":"A study of the effects in signal reconstruction from degraded Fourier transform phase","authors":"A. Nandi","doi":"10.1109/MDSP.1989.97119","DOIUrl":"https://doi.org/10.1109/MDSP.1989.97119","url":null,"abstract":"Summary form only given. The reconstruction of signals from phase only information has been treated. Although this is not possible in general, it has been shown previously that a finite duration sequence, provided its z-transform has no zeros in reciprocal pairs or on the unit circle, is uniquely specified by its Fourier transform phase within a scale factor and therefore can be reconstructed within a scale factor. The derivation assumes that the phases are known exactly; no error of any kind is involved. As all measured quantities are degraded by some kind of error, this assumption is much too restrictive when one wishes to reconstruct real signals as opposed to simulated signals. To study the effects of phase degradation on the quality of reconstructed signals, a series of experiments has been performed. The effects of additive random noise and of quantization noise in the phase samples have been investigated along with the effects of uniform and nonuniform sampling of the Fourier phase function. It has been found that the quality of a reconstructed signal is dependent on the choice of the set of frequencies used for sampling the Fourier phase function.<<ETX>>","PeriodicalId":340681,"journal":{"name":"Sixth Multidimensional Signal Processing Workshop,","volume":"415 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1989-09-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117292734","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Summary form only given. Fast algorithms for computing the linear least-squares estimate of a multidimensional random field from noisy observations inside a circle (2-D) or sphere (3-D) have been derived. The double Radon transform of the random field covariance is assumed to have to Toeplitz-plus-Hankel structure. The algorithms can be viewed as general split Levinson and Schur algorithms, since they exploit this structure in the same way that their one-dimensional counterparts exploit the Toeplitz structure of the covariance of a stationary random process. The algorithm are easily parallelizable, and they are recursive in increasing radius of the hypersphere of observations. A discrete form of the problem and a discrete algorithm for solving it was included. Numerical results on the performance of the algorithm have been obtained. A procedure for estimating a covariance of the desired form from a sample function of a random field (i.e. a multidimensional 'Toeplitzation plus Hankelization') and a one-dimensional discrete algorithm for arbitrary Toeplitz-plus-Hankel systems of equations.<>
{"title":"Generalized Levinson and Schur algorithms for multi-dimensional random field estimation problems","authors":"A. Yagle","doi":"10.1109/MDSP.1989.97084","DOIUrl":"https://doi.org/10.1109/MDSP.1989.97084","url":null,"abstract":"Summary form only given. Fast algorithms for computing the linear least-squares estimate of a multidimensional random field from noisy observations inside a circle (2-D) or sphere (3-D) have been derived. The double Radon transform of the random field covariance is assumed to have to Toeplitz-plus-Hankel structure. The algorithms can be viewed as general split Levinson and Schur algorithms, since they exploit this structure in the same way that their one-dimensional counterparts exploit the Toeplitz structure of the covariance of a stationary random process. The algorithm are easily parallelizable, and they are recursive in increasing radius of the hypersphere of observations. A discrete form of the problem and a discrete algorithm for solving it was included. Numerical results on the performance of the algorithm have been obtained. A procedure for estimating a covariance of the desired form from a sample function of a random field (i.e. a multidimensional 'Toeplitzation plus Hankelization') and a one-dimensional discrete algorithm for arbitrary Toeplitz-plus-Hankel systems of equations.<<ETX>>","PeriodicalId":340681,"journal":{"name":"Sixth Multidimensional Signal Processing Workshop,","volume":"28 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1989-09-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132661696","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Summary form only given. The quality of the image displayed on the workstation monitor is inferior to that seen through the camera lens and it has been determined that there is no theoretical reason for this phenomenon. From the Nyquist sampling theorem and Fraunhofer diffraction theory it is known that the sampling frequency should be more than twice the highest frequency in the image plane of a camera lens. In microscopy, for a numerical aperture (NA) of 1.3 and a wavelength of 500 nm a sampling density of approximately 100 nm/pixel or a sampling frequency of 10 pixels/ mu m is required. If this condition is met, then the digitized information stored in the computer memory is sufficient to reconstruct the continuous image as seen through the lens. This reconstruction problem has been analyzed in detail, and it has been determined that instead of using the standard reconstruction procedure based on sinc functions, it is possible to reconstruct exactly a continuous chromosome image using a finite number of samples. This leads to the possibility of high-density resampling of the image to provide displays of arbitrarily high quality.<>
{"title":"Exact reconstruction of sampled images","authors":"I. Young, P. Verbeek, H. Netten","doi":"10.1109/MDSP.1989.97127","DOIUrl":"https://doi.org/10.1109/MDSP.1989.97127","url":null,"abstract":"Summary form only given. The quality of the image displayed on the workstation monitor is inferior to that seen through the camera lens and it has been determined that there is no theoretical reason for this phenomenon. From the Nyquist sampling theorem and Fraunhofer diffraction theory it is known that the sampling frequency should be more than twice the highest frequency in the image plane of a camera lens. In microscopy, for a numerical aperture (NA) of 1.3 and a wavelength of 500 nm a sampling density of approximately 100 nm/pixel or a sampling frequency of 10 pixels/ mu m is required. If this condition is met, then the digitized information stored in the computer memory is sufficient to reconstruct the continuous image as seen through the lens. This reconstruction problem has been analyzed in detail, and it has been determined that instead of using the standard reconstruction procedure based on sinc functions, it is possible to reconstruct exactly a continuous chromosome image using a finite number of samples. This leads to the possibility of high-density resampling of the image to provide displays of arbitrarily high quality.<<ETX>>","PeriodicalId":340681,"journal":{"name":"Sixth Multidimensional Signal Processing Workshop,","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1989-09-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132759153","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Summary form only given. A two-stage coding system for reducing the high data rate is reported. Three-dimensional subsampling is used to create the baseband of the interlaced source signal. A motion-adaptive filter structure adjusts the three-dimensional spectrum of the television signal to some reduced region supported by the quincunx sampling pattern. By interpolating the subsampled signal (main signal) in the transmitter, calculating the difference from the original, and subsampling the difference signal again, the error signal results. Both signals have to be coded to achieve a complete bit rate of less than 140 Mb/s. Transform coding is applied to the main signal. A modified threshold-coding scheme is applied to the DCT coefficients. In order to avoid artifacts in the reconstructed (moving) sequence, threshold, normalization, and quantization of the coefficients are adapted to the local picture content. In regions where the spatial resolution has been reduced by the filtering process, the error signal is coded, transmitted and added to th main signal to improve the picture quality. Tree coding has been investigated for this purpose. Simulations of the algorithms have been performed with TV and HDTV sequences.<>
{"title":"3D-frequency splitting and coding of HDTV signals","authors":"G. Schamel","doi":"10.1109/MDSP.1989.97136","DOIUrl":"https://doi.org/10.1109/MDSP.1989.97136","url":null,"abstract":"Summary form only given. A two-stage coding system for reducing the high data rate is reported. Three-dimensional subsampling is used to create the baseband of the interlaced source signal. A motion-adaptive filter structure adjusts the three-dimensional spectrum of the television signal to some reduced region supported by the quincunx sampling pattern. By interpolating the subsampled signal (main signal) in the transmitter, calculating the difference from the original, and subsampling the difference signal again, the error signal results. Both signals have to be coded to achieve a complete bit rate of less than 140 Mb/s. Transform coding is applied to the main signal. A modified threshold-coding scheme is applied to the DCT coefficients. In order to avoid artifacts in the reconstructed (moving) sequence, threshold, normalization, and quantization of the coefficients are adapted to the local picture content. In regions where the spatial resolution has been reduced by the filtering process, the error signal is coded, transmitted and added to th main signal to improve the picture quality. Tree coding has been investigated for this purpose. Simulations of the algorithms have been performed with TV and HDTV sequences.<<ETX>>","PeriodicalId":340681,"journal":{"name":"Sixth Multidimensional Signal Processing Workshop,","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1989-09-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131912377","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Summary form only given, as follows. Most array processing schemes rely on the use of a signal replica correlated with the observations to detect and localize targets of interest. Matched field processors make use of signal replicas that are accurately tuned to available environmental knowledge. When knowledge about the array system, such as sensor positions, or environmental parameters, such as sound speed, which are used to form the matched field signal replica, is imprecise, this causes a mismatch between the replica and the actual signal and the performance of the processor may be seriously degraded. Analytic methods for predicting the sensitivity of matched field processors to replica mismatch are developed. Bounds on the overall effect of mismatch are also developed. The use of these methods is illustrated through discussion of an example. Matched-field array processing methods can, in many situations, significantly improve target detection and localization performance. This work provides one of the only analytical tools that can be used to assess the performance of such processors in the context of real-world system limitations.<>
{"title":"Methods for predicting the sensitivity of matched field processors to replica mismatch","authors":"D. Gingras","doi":"10.1109/MDSP.1989.97041","DOIUrl":"https://doi.org/10.1109/MDSP.1989.97041","url":null,"abstract":"Summary form only given, as follows. Most array processing schemes rely on the use of a signal replica correlated with the observations to detect and localize targets of interest. Matched field processors make use of signal replicas that are accurately tuned to available environmental knowledge. When knowledge about the array system, such as sensor positions, or environmental parameters, such as sound speed, which are used to form the matched field signal replica, is imprecise, this causes a mismatch between the replica and the actual signal and the performance of the processor may be seriously degraded. Analytic methods for predicting the sensitivity of matched field processors to replica mismatch are developed. Bounds on the overall effect of mismatch are also developed. The use of these methods is illustrated through discussion of an example. Matched-field array processing methods can, in many situations, significantly improve target detection and localization performance. This work provides one of the only analytical tools that can be used to assess the performance of such processors in the context of real-world system limitations.<<ETX>>","PeriodicalId":340681,"journal":{"name":"Sixth Multidimensional Signal Processing Workshop,","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1989-09-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133786754","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Summary form only given, as follows. A multiresolution edge detection can be performed with a wavelet transform. Indeed, for some particular wavelets, the wavelet transform of an image provides the local gradient of the image at different resolutions. A multiresolution edge detection is therefore equivalent to a detection of local extrema in the image wavelet transform (local extrema of the image gradient). It is shown that one can build a complete image representation by recording the value and the position of these local extrema on a dyadic sequence of resolutions: 1/2, 1/4, 1/8 etc. An iterative procedure that reconstructs the image from these local extrema is described. The algorithm is based on the reproducing kernal of a wavelet transform; it is numerically stable. This reconstruction shows that an image can be coded from the edges which appear on a dyadic sequence of resolutions, without losing any information. Such an adaptive coding is useful for pattern recognition but also for data compression. Indeed, the edges of an image can be efficiently coded into chains with predictive techniques.<>
{"title":"Image coding from the wavelet transform extrema","authors":"S. Mallat, N. Treil, S. Zhong","doi":"10.1109/MDSP.1989.97055","DOIUrl":"https://doi.org/10.1109/MDSP.1989.97055","url":null,"abstract":"Summary form only given, as follows. A multiresolution edge detection can be performed with a wavelet transform. Indeed, for some particular wavelets, the wavelet transform of an image provides the local gradient of the image at different resolutions. A multiresolution edge detection is therefore equivalent to a detection of local extrema in the image wavelet transform (local extrema of the image gradient). It is shown that one can build a complete image representation by recording the value and the position of these local extrema on a dyadic sequence of resolutions: 1/2, 1/4, 1/8 etc. An iterative procedure that reconstructs the image from these local extrema is described. The algorithm is based on the reproducing kernal of a wavelet transform; it is numerically stable. This reconstruction shows that an image can be coded from the edges which appear on a dyadic sequence of resolutions, without losing any information. Such an adaptive coding is useful for pattern recognition but also for data compression. Indeed, the edges of an image can be efficiently coded into chains with predictive techniques.<<ETX>>","PeriodicalId":340681,"journal":{"name":"Sixth Multidimensional Signal Processing Workshop,","volume":"36 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1989-09-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115202984","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Summary form only given. The discrete Fourier transform has been implemented on a single-instruction, multiple-data (SIMD) machine. The implementation demonstrates how an algorithm that is unsuited for use on a sequential machine can be very effective in a parallel machine. The SIMD machine is based on the Geometric Arithmetic Parallel Processor (GAPP). The Bluestein chirp algorithm, a variation of the chirp-Z algorithm, is the key to parallelizing the Fourier transform. When the chirp-Z is adapted to the parallel architecture of the GAPP array, the transform is reduced to O(N) operations as compared to O(N*N log N) on sequential machines. The GAPP array used to implement this algorithm is a 108*384 array. Each processing element is a one-bit serial ALU with 128 bits of RAM. Each processor is connected to its four nearest neighbors (north, south, east, and west) in a mesh configuration.<>
{"title":"Implementation of an N*N Fourier transform in order N instructions on a SIMD array","authors":"A. Chang, J. Selvage, A. Forman, P. Walker","doi":"10.1109/MDSP.1989.97097","DOIUrl":"https://doi.org/10.1109/MDSP.1989.97097","url":null,"abstract":"Summary form only given. The discrete Fourier transform has been implemented on a single-instruction, multiple-data (SIMD) machine. The implementation demonstrates how an algorithm that is unsuited for use on a sequential machine can be very effective in a parallel machine. The SIMD machine is based on the Geometric Arithmetic Parallel Processor (GAPP). The Bluestein chirp algorithm, a variation of the chirp-Z algorithm, is the key to parallelizing the Fourier transform. When the chirp-Z is adapted to the parallel architecture of the GAPP array, the transform is reduced to O(N) operations as compared to O(N*N log N) on sequential machines. The GAPP array used to implement this algorithm is a 108*384 array. Each processing element is a one-bit serial ALU with 128 bits of RAM. Each processor is connected to its four nearest neighbors (north, south, east, and west) in a mesh configuration.<<ETX>>","PeriodicalId":340681,"journal":{"name":"Sixth Multidimensional Signal Processing Workshop,","volume":"15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1989-09-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115469043","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Summary form only given. The problem of learning the parameters needed for image restoration from the given noisy and blurred image has been addressed. The asymptotically optimal approach of finding the maximum-likelihood estimate of the parameters and then using this value of the parameter to construct the restoration filter has been taken. One way to do this is to iteratively solve the nonlinear problem of maximizing the a posteriori probability of the image given the blurred observations and also the unknown parameters. The ellipsoidal algorithm and the expectation-maximization (EM) algorithm have been used for this purpose. An experimental comparison of these two methods for parametrically restoring images when the parameters are not known a priori has been made.<>
{"title":"An experimental comparison of the EM algorithm versus general optimization for combined image identification and restoration","authors":"J. Woods, S. Rastogi","doi":"10.1109/MDSP.1989.97104","DOIUrl":"https://doi.org/10.1109/MDSP.1989.97104","url":null,"abstract":"Summary form only given. The problem of learning the parameters needed for image restoration from the given noisy and blurred image has been addressed. The asymptotically optimal approach of finding the maximum-likelihood estimate of the parameters and then using this value of the parameter to construct the restoration filter has been taken. One way to do this is to iteratively solve the nonlinear problem of maximizing the a posteriori probability of the image given the blurred observations and also the unknown parameters. The ellipsoidal algorithm and the expectation-maximization (EM) algorithm have been used for this purpose. An experimental comparison of these two methods for parametrically restoring images when the parameters are not known a priori has been made.<<ETX>>","PeriodicalId":340681,"journal":{"name":"Sixth Multidimensional Signal Processing Workshop,","volume":"98 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1989-09-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115748079","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：