Segmentation of layers in retinal images obtained by optical coherence tomography(OCT)has become an important clinical tool to diagnose ophthalmic diseases.However,due to the sus-ceptibility to speckle noise and shado...Segmentation of layers in retinal images obtained by optical coherence tomography(OCT)has become an important clinical tool to diagnose ophthalmic diseases.However,due to the sus-ceptibility to speckle noise and shadow of blood vessels etc.,the layer segmentation technology based on a single image still fail to reach a satisfactory level.We propose a combination method of structure interpolation and lateral mean filtering(SI-LMF)to improve the signal-to-noise ratio based on one retinal image.Before performing one-dimensional lateral mean filtering to remove noise,structure interpolation was operated to eliminate thickness fluctuations.Then,we used boundary growth method to identify boundaries.Compared with existing segmentations,the method proposed in this paper requires less data and avoids the influence of microsaccade.The automatic segmentation method was verified on the spectral domain OCT volume images obtained from four normal objects,which successfully identified the boundaries of 10 physio-logical layers,consistent with the results based on the manual determination.展开更多
Life science research aims to continuously improve the quality and standard of human life. One of the major challenges in this area is to maintain food safety and security. A number of image processing techniques have...Life science research aims to continuously improve the quality and standard of human life. One of the major challenges in this area is to maintain food safety and security. A number of image processing techniques have been used to investigate the quality of food products. In this paper, we propose a new algorithm to effectively segment connected grains so that each of them can be inspected in a later processing stage. One family of the existing segmentation methods is based on the idea of watersheding, and it has shown promising results in practice. However, due to the over-segmentation issue, this technique has experienced poor performance in various applications, such as inhomogeneous background and connected targets. To solve this problem, we present a combination of two classical techniques to handle this issue. In the first step, a mean shift filter is used to eliminate the inhomogeneous background,where entropy is used to be a converging criterion. Secondly, a color gradient algorithm is used in order to detect the most significant edges, and a marked watershed transform is applied to segment cluttered objects out of the previous processing stages. The proposed framework is capable of compromising among execution time, usability, efficiency and segmentation outcome in analyzing ring die pellets. The experimental results demonstrate that the proposed approach is effectiveness and robust.展开更多
Underwater acoustic signal processing is one of the research hotspots in underwater acoustics.Noise reduction of underwater acoustic signals is the key to underwater acoustic signal processing.Owing to the complexity ...Underwater acoustic signal processing is one of the research hotspots in underwater acoustics.Noise reduction of underwater acoustic signals is the key to underwater acoustic signal processing.Owing to the complexity of marine environment and the particularity of underwater acoustic channel,noise reduction of underwater acoustic signals has always been a difficult challenge in the field of underwater acoustic signal processing.In order to solve the dilemma,we proposed a novel noise reduction technique for underwater acoustic signals based on complete ensemble empirical mode decomposition with adaptive noise(CEEMDAN),minimum mean square variance criterion(MMSVC) and least mean square adaptive filter(LMSAF).This noise reduction technique,named CEEMDAN-MMSVC-LMSAF,has three main advantages:(i) as an improved algorithm of empirical mode decomposition(EMD) and ensemble EMD(EEMD),CEEMDAN can better suppress mode mixing,and can avoid selecting the number of decomposition in variational mode decomposition(VMD);(ii) MMSVC can identify noisy intrinsic mode function(IMF),and can avoid selecting thresholds of different permutation entropies;(iii) for noise reduction of noisy IMFs,LMSAF overcomes the selection of deco mposition number and basis function for wavelet noise reduction.Firstly,CEEMDAN decomposes the original signal into IMFs,which can be divided into noisy IMFs and real IMFs.Then,MMSVC and LMSAF are used to detect identify noisy IMFs and remove noise components from noisy IMFs.Finally,both denoised noisy IMFs and real IMFs are reconstructed and the final denoised signal is obtained.Compared with other noise reduction techniques,the validity of CEEMDAN-MMSVC-LMSAF can be proved by the analysis of simulation signals and real underwater acoustic signals,which has the better noise reduction effect and has practical application value.CEEMDAN-MMSVC-LMSAF also provides a reliable basis for the detection,feature extraction,classification and recognition of underwater acoustic signals.展开更多
Background Image denoising is an important topic in the digital image processing field.This study theoretically investigates the validity of the classical nonlocal mean filter(NLM)for removing Gaussian noise from a no...Background Image denoising is an important topic in the digital image processing field.This study theoretically investigates the validity of the classical nonlocal mean filter(NLM)for removing Gaussian noise from a novel statistical perspective.Method By considering the restored image as an estimator of the clear image from a statistical perspective,we gradually analyze the unbiasedness and effectiveness of the restored value obtained by the NLM filter.Subsequently,we propose an improved NLM algorithm called the clustering-based NLM filter that is derived from the conditions obtained through the theoretical analysis.The proposed filter attempts to restore an ideal value using the approximately constant intensities obtained by the image clustering process.In this study,we adopt a mixed probability model on a prefiltered image to generate an estimator of the ideal clustered components.Result The experiment yields improved peak signal-to-noise ratio values and visual results upon the removal of Gaussian noise.Conclusion However,the considerable practical performance of our filter demonstrates that our method is theoretically acceptable as it can effectively estimate ideal images.展开更多
A novel multiresolution pyramidal edge detector, based on adaptive weighted fuzzy mean(AWFM)filtering and fuzzy linking model, is presented in this paper. The algorithm first constructs a pyramidal structure by repeti...A novel multiresolution pyramidal edge detector, based on adaptive weighted fuzzy mean(AWFM)filtering and fuzzy linking model, is presented in this paper. The algorithm first constructs a pyramidal structure by repetitive AWFM filtering and subsampling of original image. Then it utilizes multiple heuristic linking criteria between the edge nodes of two adjacent levels and considers the linkage as a fuzzy model, which is trained offline. Through this fuzzy linking model, the boundaries detected at coarse resolution are propagated and refined to the bottom level from the coarse-to fine edge detection. The validation experiment results demonstrate that the proposed approach has superior performance compared with standard fixed resolution detector andprevious multiresolution approach, especially in impulse noise environment.展开更多
To position personnel in mines, the study discussed in this paper built on the tunnel personnel positioning method on the basis of both TOA and location-finger print(LFP) positioning. Given non-line of sight(NLOS) tim...To position personnel in mines, the study discussed in this paper built on the tunnel personnel positioning method on the basis of both TOA and location-finger print(LFP) positioning. Given non-line of sight(NLOS) time delay in signal transmission caused by facilities and equipment shielding in tunnels and TOA measurement errors in both LFP database data and real-time data, this paper puts forth a database data de-noising algorithm based on distance threshold limitation and modified mean filtering(MMF), as well as a real-time data suppression algorithm based on speed threshold limitation and MMF.On this basis, a nearest neighboring data matching algorithm based on historical location and the speed threshold limitation is used to estimate personnel location and realize accurate personnel positioning.The results from both simulation and the experiment suggest that: compared with the basic LFP positioning method and the method that only suppresses real-time data error, the tunnel personnel positioning methods based on TOA and modified LFP positioning permits effectively eliminating error in TOA measurement, making the measured data close to the true positional data, and dropping the positioning error:the maximal positioning error in measurements from experiment drops by 9 and 3 m, respectively, and the positioning accuracy of 3 m is achievable in the condition used in the experiment.展开更多
Based on the continuous development of motion capture technology for ordinary video images, unmarked optical motion capture has become the fastest human posture recognition technology. Compared with other technical pr...Based on the continuous development of motion capture technology for ordinary video images, unmarked optical motion capture has become the fastest human posture recognition technology. Compared with other technical products, Google’s 3D human body recognition framework—Mediapipe is the most mature representative in this field. However, Mediapipe also has many defects in the detection of 3D human posture. In this paper, firstly, to solve the problem of inaccurate detection of human posture by Mediapipe, the accuracy of 2D human posture detection is improved through the speed threshold correction method for each joint;According to the problem that the monocular camera can not detect the depth Z value in the human posture data accurately, the Z value of the joint point is corrected for the human tilt angle through statistics;Then, according to the inaccurate recognition of Z value caused by large body posture, the accurate correction of Z value of human posture under different body posture is realized by normalizing the simulation proportion of each body limb;Finally, in order to solve the problem of jitter, lag problem and periodic noise in multiple frames caused by the speed change of human joints, one euro filtering and mean filtering of joint data are carried out. This paper verifies that the accuracy of 3D human posture detection based on the improved Mediapipe is more than 90% through the multi-pose recognition test for people of different heights, weights, ages and gender.展开更多
The principle of the support vector regression machine（SVR） is first analysed. Then the new data-dependent kernel function is constructed from information geometry perspective. The current waveforms change regularly...The principle of the support vector regression machine（SVR） is first analysed. Then the new data-dependent kernel function is constructed from information geometry perspective. The current waveforms change regularly in accordance with the different horizontal offset when the rotational frequency of the high speed rotational arc sensor is in the range from 15 Hz to 30 Hz. The welding current data is pretreated by wavelet filtering, mean filtering and normalization treatment. The SVR model is constructed by making use of the evolvement laws, the decision function can be achieved by training the SVR and the seam offset can be identified. The experimental results show that the precision of the offset identification can be greatly improved by modifying the SVR and applying mean filteringfrom the longitudinal direction.展开更多
It is quite often that the theoretic model used in the Kalman filtering may not be sufficiently accurate for practical applications,due to the fact that the covariances of noises are not exactly known.Our previous wor...It is quite often that the theoretic model used in the Kalman filtering may not be sufficiently accurate for practical applications,due to the fact that the covariances of noises are not exactly known.Our previous work reveals that in such scenario the filter calculated mean square errors(FMSE)and the true mean square errors(TMSE)become inconsistent,while FMSE and TMSE are consistent in the Kalman filter with accurate models.This can lead to low credibility of state estimation regardless of using Kalman filters or adaptive Kalman filters.Obviously,it is important to study the inconsistency issue since it is vital to understand the quantitative influence induced by the inaccurate models.Aiming at this,the concept of credibility is adopted to discuss the inconsistency problem in this paper.In order to formulate the degree of the credibility,a trust factor is constructed based on the FMSE and the TMSE.However,the trust factor can not be directly computed since the TMSE cannot be found for practical applications.Based on the definition of trust factor,the estimation of the trust factor is successfully modified to online estimation of the TMSE.More importantly,a necessary and sufficient condition is found,which turns out to be the basis for better design of Kalman filters with high performance.Accordingly,beyond trust factor estimation with Sage-Husa technique(TFE-SHT),three novel trust factor estimation methods,which are directly numerical solving method(TFE-DNS),the particle swarm optimization method(PSO)and expectation maximization-particle swarm optimization method(EM-PSO)are proposed.The analysis and simulation results both show that the proposed TFE-DNS is better than the TFE-SHT for the case of single unknown noise covariance.Meanwhile,the proposed EMPSO performs completely better than the EM and PSO on the estimation of the credibility degree and state when both noise covariances should be estimated online.展开更多
The spatial diversity of distributed network demands the individual filter to accommodate the topology of interference environment. In this paper, a type of distributed adaptive beamformer is proposed to mitigate inte...The spatial diversity of distributed network demands the individual filter to accommodate the topology of interference environment. In this paper, a type of distributed adaptive beamformer is proposed to mitigate interference over coordinated antenna arrays network. The proposed approach is formulated as generalized sidelobe canceller （GSC） structure to facilitate the convex combination of neighboring nodes＇ weights, and then it is solved by unconstrained least mean square （LMS） algorithm due to simplicity. Numerical results show that the robustness and convergence rate of antenna arrays network can be significantly improved in strong interference scenario. And they also clearly illustrate that mixing vector is optimized adaptively and adjusted according to the spatial diversity of the distributed nodes which are placed in different power of received signals to interference ratio （SIR） environments.展开更多
Aeromagnetic interference could not be compensated effectively if the precision of parameters which are solved by the aircraft magnetic field model is low. In order to improve the compensation effect under this condit...Aeromagnetic interference could not be compensated effectively if the precision of parameters which are solved by the aircraft magnetic field model is low. In order to improve the compensation effect under this condition, a method based on small signal model and least mean square（LMS） algorithm is proposed. According to the method, the initial values of adaptive filter＇s weight vector are calculated with the solved model parameters through small signal model at first,then the small amount of direction cosine and its derivative are set as the input of the filter, and the small amount of the interference is set as the filter＇s expected vector. After that, the aircraft magnetic interference is compensated by LMS algorithm. Finally, the method is verified by simulation and experiment. The result shows that the compensation effect can be improved obviously by the LMS algorithm when original solved parameters have low precision. The method can further improve the compensation effect even if the solved parameters have high precision.展开更多
Dynamic performance of a reactive power compensation （RPC） system for the international thermonuclear experimental reactor （ITER） power supply is presented. Static var compensators （SVCs） are adopted to mitigate...Dynamic performance of a reactive power compensation （RPC） system for the international thermonuclear experimental reactor （ITER） power supply is presented. Static var compensators （SVCs） are adopted to mitigate voltage fluctuation and reduce the reactive power down to a level acceptable for the French/European 400 kV grid. A voltage feedback and load power feedforward controller for SVC is proposed, with the feedforward loop intended to guarantee short response time and the feedback loop ensuring good dynamics and steady characteristics of SVC. A mean filter was chosen to measure the control signals to improve the dynamic response. The dynamic performance of the SVC is verified by simulations using PSCAD/EMTDC codes.展开更多
Super-resolution techniques are used to reconstruct an image with a high resolution from one or more low-resolution image(s).In this paper,we proposed a single image super-resolution algorithm.It uses the nonlocal mea...Super-resolution techniques are used to reconstruct an image with a high resolution from one or more low-resolution image(s).In this paper,we proposed a single image super-resolution algorithm.It uses the nonlocal mean filter as a prior step to produce a denoised image.The proposed algorithm is based on curvelet transform.It converts the denoised image into low and high frequencies(sub-bands).Then we applied a multi-dimensional interpolation called Lancozos interpolation over both sub-bands.In parallel,we applied sparse representation with over complete dictionary for the denoised image.The proposed algorithm then combines the dictionary learning in the sparse representation and the interpolated sub-bands using inverse curvelet transform to have an image with a higher resolution.The experimental results of the proposed super-resolution algorithm show superior performance and obviously better-recovering images with enhanced edges.The comparison study shows that the proposed super-resolution algorithm outperforms the state-of-the-art.The mean absolute error is 0.021±0.008 and the structural similarity index measure is 0.89±0.08.展开更多
Fabric defect detection has been an indispensable and important link in fabric production,many studies on the development of vision based automated inspection techniques have been reported.The main drawback of existin...Fabric defect detection has been an indispensable and important link in fabric production,many studies on the development of vision based automated inspection techniques have been reported.The main drawback of existing methods is that they can only inspect a particular type of fabric pattern in controlled environment.Recently,nonlocal self-similarity(NSS)based method is used for fabric defect detection.This method achieves good defect detection performance for small defects with uneven illumination,the disadvantage of NNS based method is poor for detecting linear defects.Based on this reason,we improve NSS based defect detection method by introducing a gray density function,namely an enhanced NSS(ENSS)based defect detection method.Meanwhile,mean filter is applied to smooth images and suppress noise.Experimental results prove the validity and feasibility of the proposed NLRA algorithm.展开更多
Rafael C. Gnzalez has mentioned an algorithm on adaptive local noise elimination filter in the book named Digital Image Processing. This paper points out the algorithm＇s deficiency and presents an improved harmonic m...Rafael C. Gnzalez has mentioned an algorithm on adaptive local noise elimination filter in the book named Digital Image Processing. This paper points out the algorithm＇s deficiency and presents an improved harmonic mean filter algorithm which makes mean square error emse cutting quarter but SNR, SNPm and PSNR increasing a tenth more than original algorithm. This filter algorithm is verified to be effective by simulation experiment.展开更多
For segmenting cerebral blood vessels from the time-of-flight magnetic resonance angiography （TOF-MRA） images accurately, we propose a parallel segmentation algorithm based on statistical model with Markov random fi...For segmenting cerebral blood vessels from the time-of-flight magnetic resonance angiography （TOF-MRA） images accurately, we propose a parallel segmentation algorithm based on statistical model with Markov random field （MRF）. Firstly, we improve traditional non-local means filter with patch-based Fourier transformation to preprocess the TOF-MRA images. In this step, we mainly utilize the sparseness and self-similarity of the MRA brain images sequence. Secondly, we add the MRF information to the finite mixture mode （FMM） to fit the intensity distribution of medical images. We make use of the MRF in image sequence to estimate the proportion of cerebral tissues. Finally, we choose the particle swarm optimization （PSO） algorithm to parallelize the parameter estimation of FMM. A large number of experiments verify the high accuracy and robustness of our approach especially for narrow vessels. The work will offer significant assistance for physicians on the prevention and diagnosis of cerebrovascular diseases.展开更多
In this study, hybrid computational frameworks are developed for active noise control（ANC） systems using an evolutionary computing technique based on genetic algorithms（GAs） and interior-point method（IPM）, follo...In this study, hybrid computational frameworks are developed for active noise control（ANC） systems using an evolutionary computing technique based on genetic algorithms（GAs） and interior-point method（IPM）, following an integrated approach, GA-IPM. Standard ANC systems are usually implemented with the filtered extended least mean square algorithm for optimization of coefficients for the linear finite-impulse response filter, but are likely to become trapped in local minima（LM）. This issue is addressed with the proposed GA-IPM computing approach which is considerably less prone to the LM problem. Also, there is no requirement to identify a secondary path for the ANC system used in the scheme. The design method is evaluated using an ANC model of a headset with sinusoidal, random, and complex random noise interferences under several scenarios based on linear and nonlinear primary and secondary paths. The accuracy and convergence of the proposed scheme are validated based on the results of statistical analysis of a large number of independent runs of the algorithm.展开更多
基金This work was supported in part by National Natural Science Foundation of China(61771119 and 61901100)Hebei Provincial Natural Science Foundation of China(H2018501087 and H2019501010)Fundamental Research Funds for the Central Universities(N182304008).
文摘Segmentation of layers in retinal images obtained by optical coherence tomography(OCT)has become an important clinical tool to diagnose ophthalmic diseases.However,due to the sus-ceptibility to speckle noise and shadow of blood vessels etc.,the layer segmentation technology based on a single image still fail to reach a satisfactory level.We propose a combination method of structure interpolation and lateral mean filtering(SI-LMF)to improve the signal-to-noise ratio based on one retinal image.Before performing one-dimensional lateral mean filtering to remove noise,structure interpolation was operated to eliminate thickness fluctuations.Then,we used boundary growth method to identify boundaries.Compared with existing segmentations,the method proposed in this paper requires less data and avoids the influence of microsaccade.The automatic segmentation method was verified on the spectral domain OCT volume images obtained from four normal objects,which successfully identified the boundaries of 10 physio-logical layers,consistent with the results based on the manual determination.
基金supported by National Key Scientific Apparatus Development of Special Item of China(No.2012YQ15008703)Nantong Research Program of Application Foundation(No.BK2012030)Key Project of Science and Technology Commission of Shanghai Municipality(No.14JC1402200)
文摘Life science research aims to continuously improve the quality and standard of human life. One of the major challenges in this area is to maintain food safety and security. A number of image processing techniques have been used to investigate the quality of food products. In this paper, we propose a new algorithm to effectively segment connected grains so that each of them can be inspected in a later processing stage. One family of the existing segmentation methods is based on the idea of watersheding, and it has shown promising results in practice. However, due to the over-segmentation issue, this technique has experienced poor performance in various applications, such as inhomogeneous background and connected targets. To solve this problem, we present a combination of two classical techniques to handle this issue. In the first step, a mean shift filter is used to eliminate the inhomogeneous background,where entropy is used to be a converging criterion. Secondly, a color gradient algorithm is used in order to detect the most significant edges, and a marked watershed transform is applied to segment cluttered objects out of the previous processing stages. The proposed framework is capable of compromising among execution time, usability, efficiency and segmentation outcome in analyzing ring die pellets. The experimental results demonstrate that the proposed approach is effectiveness and robust.
基金The authors gratefully acknowledge the support of the National Natural Science Foundation of China(No.11574250).
文摘Underwater acoustic signal processing is one of the research hotspots in underwater acoustics.Noise reduction of underwater acoustic signals is the key to underwater acoustic signal processing.Owing to the complexity of marine environment and the particularity of underwater acoustic channel,noise reduction of underwater acoustic signals has always been a difficult challenge in the field of underwater acoustic signal processing.In order to solve the dilemma,we proposed a novel noise reduction technique for underwater acoustic signals based on complete ensemble empirical mode decomposition with adaptive noise(CEEMDAN),minimum mean square variance criterion(MMSVC) and least mean square adaptive filter(LMSAF).This noise reduction technique,named CEEMDAN-MMSVC-LMSAF,has three main advantages:(i) as an improved algorithm of empirical mode decomposition(EMD) and ensemble EMD(EEMD),CEEMDAN can better suppress mode mixing,and can avoid selecting the number of decomposition in variational mode decomposition(VMD);(ii) MMSVC can identify noisy intrinsic mode function(IMF),and can avoid selecting thresholds of different permutation entropies;(iii) for noise reduction of noisy IMFs,LMSAF overcomes the selection of deco mposition number and basis function for wavelet noise reduction.Firstly,CEEMDAN decomposes the original signal into IMFs,which can be divided into noisy IMFs and real IMFs.Then,MMSVC and LMSAF are used to detect identify noisy IMFs and remove noise components from noisy IMFs.Finally,both denoised noisy IMFs and real IMFs are reconstructed and the final denoised signal is obtained.Compared with other noise reduction techniques,the validity of CEEMDAN-MMSVC-LMSAF can be proved by the analysis of simulation signals and real underwater acoustic signals,which has the better noise reduction effect and has practical application value.CEEMDAN-MMSVC-LMSAF also provides a reliable basis for the detection,feature extraction,classification and recognition of underwater acoustic signals.
文摘Background Image denoising is an important topic in the digital image processing field.This study theoretically investigates the validity of the classical nonlocal mean filter(NLM)for removing Gaussian noise from a novel statistical perspective.Method By considering the restored image as an estimator of the clear image from a statistical perspective,we gradually analyze the unbiasedness and effectiveness of the restored value obtained by the NLM filter.Subsequently,we propose an improved NLM algorithm called the clustering-based NLM filter that is derived from the conditions obtained through the theoretical analysis.The proposed filter attempts to restore an ideal value using the approximately constant intensities obtained by the image clustering process.In this study,we adopt a mixed probability model on a prefiltered image to generate an estimator of the ideal clustered components.Result The experiment yields improved peak signal-to-noise ratio values and visual results upon the removal of Gaussian noise.Conclusion However,the considerable practical performance of our filter demonstrates that our method is theoretically acceptable as it can effectively estimate ideal images.
文摘A novel multiresolution pyramidal edge detector, based on adaptive weighted fuzzy mean(AWFM)filtering and fuzzy linking model, is presented in this paper. The algorithm first constructs a pyramidal structure by repetitive AWFM filtering and subsampling of original image. Then it utilizes multiple heuristic linking criteria between the edge nodes of two adjacent levels and considers the linkage as a fuzzy model, which is trained offline. Through this fuzzy linking model, the boundaries detected at coarse resolution are propagated and refined to the bottom level from the coarse-to fine edge detection. The validation experiment results demonstrate that the proposed approach has superior performance compared with standard fixed resolution detector andprevious multiresolution approach, especially in impulse noise environment.
基金Project supports from the National Science Foundation of China(No.51134024)the National High Technology Research and development Program of China(No.2012AA062203)are acknowledged
文摘To position personnel in mines, the study discussed in this paper built on the tunnel personnel positioning method on the basis of both TOA and location-finger print(LFP) positioning. Given non-line of sight(NLOS) time delay in signal transmission caused by facilities and equipment shielding in tunnels and TOA measurement errors in both LFP database data and real-time data, this paper puts forth a database data de-noising algorithm based on distance threshold limitation and modified mean filtering(MMF), as well as a real-time data suppression algorithm based on speed threshold limitation and MMF.On this basis, a nearest neighboring data matching algorithm based on historical location and the speed threshold limitation is used to estimate personnel location and realize accurate personnel positioning.The results from both simulation and the experiment suggest that: compared with the basic LFP positioning method and the method that only suppresses real-time data error, the tunnel personnel positioning methods based on TOA and modified LFP positioning permits effectively eliminating error in TOA measurement, making the measured data close to the true positional data, and dropping the positioning error:the maximal positioning error in measurements from experiment drops by 9 and 3 m, respectively, and the positioning accuracy of 3 m is achievable in the condition used in the experiment.
文摘Based on the continuous development of motion capture technology for ordinary video images, unmarked optical motion capture has become the fastest human posture recognition technology. Compared with other technical products, Google’s 3D human body recognition framework—Mediapipe is the most mature representative in this field. However, Mediapipe also has many defects in the detection of 3D human posture. In this paper, firstly, to solve the problem of inaccurate detection of human posture by Mediapipe, the accuracy of 2D human posture detection is improved through the speed threshold correction method for each joint;According to the problem that the monocular camera can not detect the depth Z value in the human posture data accurately, the Z value of the joint point is corrected for the human tilt angle through statistics;Then, according to the inaccurate recognition of Z value caused by large body posture, the accurate correction of Z value of human posture under different body posture is realized by normalizing the simulation proportion of each body limb;Finally, in order to solve the problem of jitter, lag problem and periodic noise in multiple frames caused by the speed change of human joints, one euro filtering and mean filtering of joint data are carried out. This paper verifies that the accuracy of 3D human posture detection based on the improved Mediapipe is more than 90% through the multi-pose recognition test for people of different heights, weights, ages and gender.
基金Supported by National Natural Science Foundation of China（ No. 50705030）.
文摘The principle of the support vector regression machine（SVR） is first analysed. Then the new data-dependent kernel function is constructed from information geometry perspective. The current waveforms change regularly in accordance with the different horizontal offset when the rotational frequency of the high speed rotational arc sensor is in the range from 15 Hz to 30 Hz. The welding current data is pretreated by wavelet filtering, mean filtering and normalization treatment. The SVR model is constructed by making use of the evolvement laws, the decision function can be achieved by training the SVR and the seam offset can be identified. The experimental results show that the precision of the offset identification can be greatly improved by modifying the SVR and applying mean filteringfrom the longitudinal direction.
基金supported by the National Natural Science Foundation of China(62033010)Aeronautical Science Foundation of China(2019460T5001)。
文摘It is quite often that the theoretic model used in the Kalman filtering may not be sufficiently accurate for practical applications,due to the fact that the covariances of noises are not exactly known.Our previous work reveals that in such scenario the filter calculated mean square errors(FMSE)and the true mean square errors(TMSE)become inconsistent,while FMSE and TMSE are consistent in the Kalman filter with accurate models.This can lead to low credibility of state estimation regardless of using Kalman filters or adaptive Kalman filters.Obviously,it is important to study the inconsistency issue since it is vital to understand the quantitative influence induced by the inaccurate models.Aiming at this,the concept of credibility is adopted to discuss the inconsistency problem in this paper.In order to formulate the degree of the credibility,a trust factor is constructed based on the FMSE and the TMSE.However,the trust factor can not be directly computed since the TMSE cannot be found for practical applications.Based on the definition of trust factor,the estimation of the trust factor is successfully modified to online estimation of the TMSE.More importantly,a necessary and sufficient condition is found,which turns out to be the basis for better design of Kalman filters with high performance.Accordingly,beyond trust factor estimation with Sage-Husa technique(TFE-SHT),three novel trust factor estimation methods,which are directly numerical solving method(TFE-DNS),the particle swarm optimization method(PSO)and expectation maximization-particle swarm optimization method(EM-PSO)are proposed.The analysis and simulation results both show that the proposed TFE-DNS is better than the TFE-SHT for the case of single unknown noise covariance.Meanwhile,the proposed EMPSO performs completely better than the EM and PSO on the estimation of the credibility degree and state when both noise covariances should be estimated online.
基金supported by National Basic Research Program of China (No. 2010CB731903)
文摘The spatial diversity of distributed network demands the individual filter to accommodate the topology of interference environment. In this paper, a type of distributed adaptive beamformer is proposed to mitigate interference over coordinated antenna arrays network. The proposed approach is formulated as generalized sidelobe canceller （GSC） structure to facilitate the convex combination of neighboring nodes＇ weights, and then it is solved by unconstrained least mean square （LMS） algorithm due to simplicity. Numerical results show that the robustness and convergence rate of antenna arrays network can be significantly improved in strong interference scenario. And they also clearly illustrate that mixing vector is optimized adaptively and adjusted according to the spatial diversity of the distributed nodes which are placed in different power of received signals to interference ratio （SIR） environments.
基金co-supported by the National Basic Research Program of China (No. 623125020103)
文摘Aeromagnetic interference could not be compensated effectively if the precision of parameters which are solved by the aircraft magnetic field model is low. In order to improve the compensation effect under this condition, a method based on small signal model and least mean square（LMS） algorithm is proposed. According to the method, the initial values of adaptive filter＇s weight vector are calculated with the solved model parameters through small signal model at first,then the small amount of direction cosine and its derivative are set as the input of the filter, and the small amount of the interference is set as the filter＇s expected vector. After that, the aircraft magnetic interference is compensated by LMS algorithm. Finally, the method is verified by simulation and experiment. The result shows that the compensation effect can be improved obviously by the LMS algorithm when original solved parameters have low precision. The method can further improve the compensation effect even if the solved parameters have high precision.
文摘Dynamic performance of a reactive power compensation （RPC） system for the international thermonuclear experimental reactor （ITER） power supply is presented. Static var compensators （SVCs） are adopted to mitigate voltage fluctuation and reduce the reactive power down to a level acceptable for the French/European 400 kV grid. A voltage feedback and load power feedforward controller for SVC is proposed, with the feedforward loop intended to guarantee short response time and the feedback loop ensuring good dynamics and steady characteristics of SVC. A mean filter was chosen to measure the control signals to improve the dynamic response. The dynamic performance of the SVC is verified by simulations using PSCAD/EMTDC codes.
文摘Super-resolution techniques are used to reconstruct an image with a high resolution from one or more low-resolution image(s).In this paper,we proposed a single image super-resolution algorithm.It uses the nonlocal mean filter as a prior step to produce a denoised image.The proposed algorithm is based on curvelet transform.It converts the denoised image into low and high frequencies(sub-bands).Then we applied a multi-dimensional interpolation called Lancozos interpolation over both sub-bands.In parallel,we applied sparse representation with over complete dictionary for the denoised image.The proposed algorithm then combines the dictionary learning in the sparse representation and the interpolated sub-bands using inverse curvelet transform to have an image with a higher resolution.The experimental results of the proposed super-resolution algorithm show superior performance and obviously better-recovering images with enhanced edges.The comparison study shows that the proposed super-resolution algorithm outperforms the state-of-the-art.The mean absolute error is 0.021±0.008 and the structural similarity index measure is 0.89±0.08.
文摘Fabric defect detection has been an indispensable and important link in fabric production,many studies on the development of vision based automated inspection techniques have been reported.The main drawback of existing methods is that they can only inspect a particular type of fabric pattern in controlled environment.Recently,nonlocal self-similarity(NSS)based method is used for fabric defect detection.This method achieves good defect detection performance for small defects with uneven illumination,the disadvantage of NNS based method is poor for detecting linear defects.Based on this reason,we improve NSS based defect detection method by introducing a gray density function,namely an enhanced NSS(ENSS)based defect detection method.Meanwhile,mean filter is applied to smooth images and suppress noise.Experimental results prove the validity and feasibility of the proposed NLRA algorithm.
基金This project is supported by National Natural Science Foundation of China （60473024） and the Natural Science Foundation of Zhejiang Province（603009）..
文摘Rafael C. Gnzalez has mentioned an algorithm on adaptive local noise elimination filter in the book named Digital Image Processing. This paper points out the algorithm＇s deficiency and presents an improved harmonic mean filter algorithm which makes mean square error emse cutting quarter but SNR, SNPm and PSNR increasing a tenth more than original algorithm. This filter algorithm is verified to be effective by simulation experiment.
基金The research is supported by the National Natural Science Foundation of China under Grant No. 61271366, and the National High Technology Research and Development 863 Program of China under Grant No. 2015AA020506.
文摘For segmenting cerebral blood vessels from the time-of-flight magnetic resonance angiography （TOF-MRA） images accurately, we propose a parallel segmentation algorithm based on statistical model with Markov random field （MRF）. Firstly, we improve traditional non-local means filter with patch-based Fourier transformation to preprocess the TOF-MRA images. In this step, we mainly utilize the sparseness and self-similarity of the MRA brain images sequence. Secondly, we add the MRF information to the finite mixture mode （FMM） to fit the intensity distribution of medical images. We make use of the MRF in image sequence to estimate the proportion of cerebral tissues. Finally, we choose the particle swarm optimization （PSO） algorithm to parallelize the parameter estimation of FMM. A large number of experiments verify the high accuracy and robustness of our approach especially for narrow vessels. The work will offer significant assistance for physicians on the prevention and diagnosis of cerebrovascular diseases.
文摘In this study, hybrid computational frameworks are developed for active noise control（ANC） systems using an evolutionary computing technique based on genetic algorithms（GAs） and interior-point method（IPM）, following an integrated approach, GA-IPM. Standard ANC systems are usually implemented with the filtered extended least mean square algorithm for optimization of coefficients for the linear finite-impulse response filter, but are likely to become trapped in local minima（LM）. This issue is addressed with the proposed GA-IPM computing approach which is considerably less prone to the LM problem. Also, there is no requirement to identify a secondary path for the ANC system used in the scheme. The design method is evaluated using an ANC model of a headset with sinusoidal, random, and complex random noise interferences under several scenarios based on linear and nonlinear primary and secondary paths. The accuracy and convergence of the proposed scheme are validated based on the results of statistical analysis of a large number of independent runs of the algorithm.
