44 results about "Morphological component analysis" patented technology

The invention discloses a face image super-resolution reconstruction method based on morphological component analysis. The method comprises the following steps: upsampling a low-resolution input image to acquire an interpolated image; acquiring a whole high-resolution face image from the interpolated image by a morphological component analysis method; downsampling the whole high-resolution face image and subtracting the whole high-resolution face image from the low-resolution input image to acquire a low-resolution residual image, and performing face detail information compensation by performing neighboring construction on an image block at each face position to acquire a high-resolution residual image; and combining the high-resolution residual image with the whole high-resolution face image to acquire the face image super-resolution result and finish resolution reconstruction. The invention also provides a method for performing face image super-resolution reconstruction and expression normalization simultaneously on the basis of morphological component analysis. By the method, detailed face detail information can be acquired, the ringing phenomenon is eliminated and high image quality is achieved.

The invention discloses a morphological component bearing failure diagnosis method based on dictionary study, and relates to the technical field of bearing failure diagnosis methods. The method comprises the steps that failure signals of an inner ring and an outer ring of a bearing are adopted as training samples, study is conducted on the training samples by applying a dictionary study algorithm, and a best dictionary space is sought; the studied dictionary replaces a fixed dictionary in a morphological component analysis method, and according to morphological difference contained of each component in the signals, separation is conducted on inner ring failure characteristics, outer ring failure characteristics and noise components in rolling bearing failure signals by utilizing the morphological component analysis method; spectral analysis is conducted on the failure characteristic components after envelope to diagnose the failure and the position of the bearing. The morphological component bearing failure diagnosis method based on the dictionary study can effectively extract failure characteristics of the outer ring and the inner ring of the rolling bearing under a strong noise environment, the principle of the method is simple, and the algorithm is easy to achieve.

The invention relates to a sea surface micro-motion target detection and characteristic extraction method based on morphological component analysis, which belongs to the technical field of radar signal treatment and detection. The method disclosed by the invention comprises the following steps: 1) sea clutter sparse representation; 2) sea clutter sparse domain inhibition; 3) the sparse representation of a micro-motion target echo signal; 4) micro-motion target signal sparse domain detection; and 5) micro-motion characteristic parameter estimation. Compared with the traditional sea surface target detection method, the sea surface micro-motion target detection and characteristic extraction method disclosed by the invention is characterized in that the morphological difference between the sea clutter and a micro-motion target echo signal composition is fully utilized, different source signals are subjected to sparse representation by different dictionaries, and the sea surface micro-motion target detection and characteristic extraction method has the capability on distinguishing the sea clutter and the micro-motion target, more signal energy is accumulated while the sea clutter is inhibited, and the signal to clutter ratio is improved. The sea surface micro-motion target detection and characteristic extraction method also has the capability on detecting the micro-motion target and estimating the micro-motion characteristic parameter in the strong sea clutter, provides a new path for sea surface weak target detection and characteristic extraction, and has a popularization and application value.

The invention discloses a sparse optimization method of suppressing strong shielding interference of seismic data. The method includes: 01, reading original seismic data, and conducting a component analysis on the data; 02, determining two sparse representation dictionaries employed by a morphological component analysis according to the morphological characteristic difference between a seismic response of a transverse heterogeneous body including a river sand body etc. related to a reservoir and a seismic response of a stable sedimentary formation in a seismic record signal, and forming an ultra-complete dictionary; 03, extracting seismic response signals of the transverse heterogeneous body including the river sand body etc. related to the reservoir from a seismic profile after suppressing overlying or underlying formation strong reflection shielding by employing a block coordinate relaxation algorithm; and 04, repeating step 03 until completion of processing of all two-dimensional measuring line data. According to the sparse optimization method of suppressing the strong shielding interference of the seismic data, the strong shielding interference can be effectively and thoroughly suppressed, and the seismic response signals of the transverse heterogeneous body including the river sand body etc. related to the reservoir can be displayed more clearly and visually.

Provided is a complicated background SAR image naval ship tail track detection method. Firstly, morphological component analysis or relative total variation regularization is carried out on an SAR image, and the image is decomposed into a texture component representing a complication sea surface and a structure component containing naval ship tail tracks; then image enhancement is carried out on the structure component and thresholding is carried out on a gradient image, and accordingly a binary image of a naval ship tail track area is formed; finally, radon transformation or Hough transformation linear detection is carried out on the enhanced naval ship tail track area, inversion is carried out in a conversion domain, and continuous tail tracks and related parameters are obtained. Structure and texture decomposition is carried out on the SAR image, and influences of speckle noise and complicated textures on naval ship tail track detection are eliminated; image enhancement is carried out on the structure component, binaryzation is carried out on the gradient image, and the naval ship tail track area is further highlighted.

The invention discloses a remote sensing image fusion method based on multi-dimensional morphologic element analysis, and belongs to the field of crossing signal processing and remote sensing image processing. The method is that the morphologic element analysis is respectively carried out for a high-resolution remote sensing image and a multi-spectrum remote sensing image under different dimension; the iterative shrinkage method is carried out to perform sparse decomposition; a target image to be fused is divided into texture components and cartoon components based on a plurality of dimensions; the cartoon component and the noise component in the high-solution image, and the texture component and the noise components in the multi-spectrum image are removed; the effective dimension texture component in the high-solution image and the cartoon component in the multi-spectrum image are remained and subjected to spare reconstruction, so as to obtain the fusion image. With the adoption of the method, the high-resolution remote sensing image and the multi-spectrum remote sensing image are effectively fused; the space resolution is improved and the spectrum distortion is reduced by being compared with the existing fusion method; in addition, the rate is greatly increased by being compared with the existing sparse reconstruction method.

The invention belongs to the field of seismic wave information processing technologies, and particularly relates to a surface wave separation method and a surface wave separation system based on morphological component analysis. The surface wave separation method based on morphological component analysis comprises the steps of: step a, making a synthetic seismogram, and constructing a surface wave separation model based on planar dictionary morphological component analysis; step b, selecting a planar non-sampling wavelet dictionary for carrying out sparse representation on surface wave signal components, and selecting a planar local discrete cosine dictionary for carrying out sparse representation on reflected wave signal components; step c, and adopting morphological component analysis under a planar dictionary for separating the surface wave signal components and the reflected wave signal components. The surface wave separation method and the surface wave separation system take both surface wave signal characteristics and reflected wave signal characteristics into account, can separate the surface wave signals and the reflected wave signals more effectively, fully utilize correlation between channels, and make the separation effect of surface wave signals better while considering that the surface wave signals are concentrated in a low-frequency high wavenumber domain under the planar non-sampling wavelet dictionary.

The invention belongs to the technical field of encryption communication and information security, and particularly relates to a digital watermark steganography method based on scrambling transformation and morphological component analysis. The method mainly comprises three processes which are sequentially the digital watermark embedding process, the communication transmission process and the digital watermark extraction process. According to the method, the morphological component analysis theory is extended to a blind source signal separation structure, and the method has the good performance effect on embedding, hiding, communication and extraction of digital watermarks and has high attack resistance and robustness for various noise attacks. According to the method, the digital watermark embedding, extraction, encryption and decryption processes are controlled through dual keys, so that transmitted data are safer.

The invention relates to a morphological component analysis (MCA)-based synthetic aperture radar (SAR) image noise suppression method. The method is technically characterized in that: an MCA method is applied to the suppression of multiplicative noises in an SAR image; and because the image generally comprises the two morphological components of edge / contour and texture, in the method, the dictionary assembly of curvelet and local discrete cosine transform (LDCT) is utilized under the uniform frame of MCA according to the good edge / contour representation characteristics of the curvelet and the high partial texture expression capability of the LDCT to fulfill the aim of suppressing the noises in the SAR image. Therefore, for a noise image, effective components of the image can be separated and the noises are kept in the remnant image by the method to fulfill the aim of suppressing the noises. Experimental results on both simulated and real SAR images show that the method has the advantage of relatively better noise suppression effect compared with the conventional wavelet-based methods and curvelet-based methods.

The invention discloses a cable coupling noise suppression method based on sparse optimization of a time-frequency domain. The method comprises the steps of 01, determining two conversion dictionaries for morphological component analysis according to morphological characteristics of an effective signal waveform and cable coupling noise, and forming an over-complete dictionary; 02, performing single track treatment on original seismic record data by using a segmented coordinate relaxation algorithm to achieve signal and noise separation, thus acquiring primarily noise-suppressed data; 03, performing high-pass filtering treatment on the primarily noise-suppressed data, subtracting the high-pass filtered noise data from the original data to acquire the effective signal data, thus eliminating wave dispersion; and repeating the steps 02-03 until the data of all tracks is processed. According to the method provided by the invention, the problem of interference of the cable coupling noise in the seismic data collected by distributed optical fibers is solved, and the cable coupling noise is thoroughly suppressed. In addition, the method provided by the invention can perform single-shot treatment, is good in adaptation, and can perform treatment in parallel, and meet industrial large-scale computing requirement.

The invention discloses an external radiation source radar wind field clutter suppression method based on morphological component analysis. A situation that a target signal and wind field clutter in the echo received by radar have different morphological characteristics is utilized, and sparse representation can be performed in corresponding transform domains so that signal separation can be performed by utilizing the morphological component analysis method. Wind field clutter can be greatly suppressed by the method. Meanwhile, the method has advantages of being great in stability and easy and convenient in calculation so that external radiation source radar target detection performance can be enhanced.

The invention belongs to the technical field of remote sensing data application and satellite image processing and particularly relates to a method for automatically classifying a satellite image scene based on morphological component analysis. The method comprises the following steps of (1) constructing a dictionary matrix through an independent meta-analysis method; (2) decomposing morphological components of a satellite image based on the morphological component analysis theory; (3) decomposing the satellite image into a texture layer and a bottom layer; (4) performing quantitative calculation on the total probability of classification of an objective image scene according to a mechanism of maximum likelihood estimation with combination of features of the texture layer and the bottom layer, and automatically classifying the satellite image scene with the quantitative calculation on the total probability of classification of objective image scenes as the basis. The method combines with a machine vision method and can be applied to a computer easily. System construction is performed more quickly and accurately through the computer than through a manual method, artificial subjective factors are eliminated, and cost is reduced. Furthermore, the method has high automatic identification accuracy and identification stability.

The invention discloses a wideband background noise suppression method for high-speed train seismic source seismic signals. Based on different frequency domain characteristics of narrow-band high-speed train seismic source seismic signals and wideband background noise, discrete cosine transform is selected to sparsely represent narrow-band high-speed train seismic source seismic signals, continuous wavelet transform is selected to sparsely represent wideband background noise, and an over-complete dictionary is constructed; a block coordinate relaxation method is then used to separate the narrow-band high-speed train seismic source seismic signals and the wideband background noise in high-speed train seismic source single-channel data, and the narrow-band high-speed train seismic source seismic signals are only kept to realize wideband background noise; and the operation is repeated until data in all channels are completed. The signal separation means such as morphological component analysis is introduced to high-speed train seismic source seismic signal processing, and wideband background noise suppression for the high-speed train seismic source seismic signals is realized.

The invention discloses a geological radar B-scan data processing method, which uses a signal processing method to process B-scan data to separate and obtain data profiles for tunnel lining anomaly and railway subgrade disease detection. Firstly, the continuous spectrum whitening frequency spectrum expansion technology is used to improve the time resolution of a signal. Then, the high-resolution data is offset and imaged by high-efficiency phase shift offset, and the lateral resolution of the profile is improved, so that the lining on the imaging profile or a reinforcing steel bar in the roadbed is focused into a point structure; finally, different morphological components such as the reinforcing steel bar and the reflective interface are separated by using a sparse separation method basedon morphological component analysis based on difference between point objects such as the reinforcing steel bar and reflective interfaces (concrete and clay layer or surrounding rock interface). Themethod helps to improve the accuracy of tunnel lining and railway subgrade detection, and provides a strong basis for tunnel and railway maintenance.

The invention discloses a meteorological radar WTC suppression method based on morphological component analysis. The method comprises the following steps: firstly, performing ground clutter suppression pretreatment on a pollution distance unit; secondly, respectively carrying out sparse representation on the preprocessed output signals in a Fourier transform domain and a short-time Fourier transform domain, and carrying out sparse separation on different morphological components by adopting a split augmented Lagrange contraction algorithm, so as to suppress WTC and recover meteorological signals on the distance unit. According to the invention, simulation experiment results show that the method can effectively inhibit ground clutters and WTC, can recover meteorological signals with high precision, and is suitable for engineering application.

The invention discloses a collected footprint pressing method based on morphological component analysis, which comprises the following steps of: 101, acquiring footprint waveform morphological characteristics in seismic records of edge-layer slices or isochronous slices of a three-dimensional seismic data body to construct two-dimensional local discrete cosine transform, and combining the two-dimensional local discrete cosine transform with two-dimensional stable wavelet transform to form an over-complete dictionary; 102, through a mode of executing original data of the seismic record signalslayer by layer, utilizing a morphological component analysis-based method to perform preliminary pressing of the collected footprints in each edge layer slice or isochronous slice; 103, determining the low intercept frequency of the two-dimensional local discrete cosine transform; 104, repeating the steps 102-103 until all slice data are processed, performing signal-to-noise separation by using amorphological component analysis-based method, suppressing and collecting footprint noise, and finally realizing suppression of the collected footprint noise in the three-dimensional seismic data body. The acquisition footprint suppression is carried out on the three-dimensional seismic data body by utilizing the method, so that the aim of improving the signal-to-noise ratio of the seismic data isfulfilled.

The present invention relates to a morphological component analysis and adaptive dictionary learning-based scene image text detection method and belongs to the digital image processing technical field. The method includes the following steps that: sample data are constructed, and two initial dictionaries, namely, a text dictionary and a background dictionary, are trained; dictionaries and parse representation coefficients corresponding to the characters and background of an image to be detected are calculated through the trained initial dictionaries, the image to be detected and an adaptive dictionary learning algorithm; a text image in the image to be detected is reconstructed through an adaptive dictionary and the parse representation coefficients corresponding to the image to be detected; the reconstructed text image is processed through using a heuristic rule, so that a candidate character area in the image to be detected can be detected out; and a final text area is framed through a rectangular box. With the method of the invention adopted, a computer can automatically understand semantic information contained in the image, and strong support can be provided for blind guiding technology, license plate recognition and vehicle positioning and tracking technology, image retrieval technology and the like.

The invention discloses a quick image segmentation method used for computer graph and image processing and based on a GPU platform and morphological component analysis. For a traditional image segmentation technology which is low in operation efficiency and poor in segmentation effect, the quick image segmentation method based on the GPU platform and the morphological component analysis is proposed so that operation time can be reduced effectively and the decomposition effect can be improved remarkably. The method includes the implementation steps that due to the fact that the advantages of GPU parallel operation are fully used, the algorithm of the morphological component analysis is achieved, the image segmentation efficiency is improved greatly, and thus quick image segmentation is achieved.

Provided is a complicated background SAR image naval ship tail track detection method. Firstly, morphological component analysis or relative total variation regularization is carried out on an SAR image, and the image is decomposed into a texture component representing a complication sea surface and a structure component containing naval ship tail tracks; then image enhancement is carried out on the structure component and thresholding is carried out on a gradient image, and accordingly a binary image of a naval ship tail track area is formed; finally, radon transformation or Hough transformation linear detection is carried out on the enhanced naval ship tail track area, inversion is carried out in a conversion domain, and continuous tail tracks and related parameters are obtained. Structure and texture decomposition is carried out on the SAR image, and influences of speckle noise and complicated textures on naval ship tail track detection are eliminated; image enhancement is carried out on the structure component, binaryzation is carried out on the gradient image, and the naval ship tail track area is further highlighted.

The invention relates to a signal robust sparse time-frequency analysis method, terminal equipment and a storage medium. The method comprises the following steps: S1, constructing a signal impact noise removal model based on morphological component analysis, sparse time-frequency analysis and frame wavelet transform; S2, for the signal impact noise removal model, calculating a texture component yT and a cartoon component yC of the signal after impact noise removal by using a forward decomposition method; and S3, according to the texture component yT and the cartoon component yC of the signal, calculating a spectrum sparse solution xi after the signal impact noise is removed by using a backward decomposition method for the signal impact noise removal model. According to the method, a morphological component analysis algorithm is introduced to decompose a signal to be processed into a low-frequency cartoon component, a high-frequency texture component and a noise component, and a stationary frame wavelet regular term is further introduced on the basis of morphological component analysis, so that the difference between a signal effective component and the noise component is fully mined, and the purposes of noise suppression and robust time-frequency analysis are achieved.

The invention discloses a multi-source fingerprint image fusion method based on convolution sparse representation. The method comprises the following steps: 1) obtaining pre-registered internal and external fingerprint images, and obtaining a sparse coefficient graph of cartoon and texture parts of the fingerprint images through employing the sparse coding of a morphological part analysis model based on convolution sparse; 2) determining a weight adding mode according to the value of the fingerprint image quality evaluation index direction determinacy, and adding the weight adding mode into cartoon and texture parts of the two source fingerprint images; and 3) respectively fusing the cartoon and texture parts of the two source fingerprint images by adopting a convolution sparse representation-based fusion method, obtaining a fusion coefficient graph of the cartoon and texture parts according to a weighted average rule, and reconstructing the fusion coefficient graph of the two parts toobtain a fused fingerprint image. According to the algorithm, more fingerprint details, textures and edge information can be reserved, regions with better quality of two source fingerprint images aresaved, and a fused fingerprint image with higher quality is obtained.

The invention discloses a visual vibration amplification method, a visual vibration detection method and a visual vibration detection system based on morphological component analysis, and belongs to the technical field of computer vision. The method comprises: acquiring a video file comprising a target object, and determining a region of interest of the target object in each frame of image in the video file; adopting a quality factor adjustable wavelet dictionary and a discrete sine and cosine dictionary to respectively represent a structure component and a texture component of each frame of video; adopting a threshold selection algorithm based on generalized Gaussian density distribution to determine thresholds of the structural components; separating an image structure component and a texture component by adopting a self-adaptive MCA; according to the separated structure components, in combination with the Euler visual angle principle, achieving amplification of micro-vibration signals, reconstructing an amplified video, meanwhile, extracting visual vibration signals for the structure components, and achieving measurement of multiple vibration frequencies.

The invention relates to an ancient book image restoration method, terminal equipment and a storage medium. The method comprises the steps of S1 constructing an ancient book image restoration model based on morphological decomposition; S2 introducing auxiliary variables Q0, Q1 and Q2 through an alternating multiplier iteration method, introducing Lagrange multiplier secondary penalty terms and secondary penalty term coefficients lambda0, lambda1 and lambda2 corresponding to the auxiliary variables Q0, Q1 and Q2, and converting solution of the formula in the step S1 into solution of FC, FT, Q0, Q1 and Q2; and S3 carrying out iterative training on the ancient book image restoration model, calculating a restored image F of the current iteration by respectively solving FC, FT, Q0, Q1 and Q2 in each iteration, further calculating whether the restored image F is true or not, and outputting the restored image F (k + 1) in the current iteration if the restored image F is true. According to the method, the ancient book image restoration model based on morphological component analysis and Lp pseudo norm is proposed, then an alternating multiplier iteration method is introduced to decompose the proposed model into a plurality of decoupled sub-problems, and then the sub-problems are solved, so that the ancient book image can be well restored.

The invention relates to a sea surface micro-motion target detection and characteristic extraction method based on morphological component analysis, which belongs to the technical field of radar signal treatment and detection. The method disclosed by the invention comprises the following steps: 1) sea clutter sparse representation; 2) sea clutter sparse domain inhibition; 3) the sparse representation of a micro-motion target echo signal; 4) micro-motion target signal sparse domain detection; and 5) micro-motion characteristic parameter estimation. Compared with the traditional sea surface target detection method, the sea surface micro-motion target detection and characteristic extraction method disclosed by the invention is characterized in that the morphological difference between the sea clutter and a micro-motion target echo signal composition is fully utilized, different source signals are subjected to sparse representation by different dictionaries, and the sea surface micro-motion target detection and characteristic extraction method has the capability on distinguishing the sea clutter and the micro-motion target, more signal energy is accumulated while the sea clutter is inhibited, and the signal to clutter ratio is improved. The sea surface micro-motion target detection and characteristic extraction method also has the capability on detecting the micro-motion target and estimating the micro-motion characteristic parameter in the strong sea clutter, provides a new path for sea surface weak target detection and characteristic extraction, and has a popularization and application value.

The invention discloses a high-speed rail seismic source seismic signal extraction method based on equidistant spectrum characteristics. The method comprises the following steps of estimating the intervals of equidistant spectrums in an amplitude spectrum of the acquired high-speed rail seismic source seismic data; then, using a morphological component analysis method, on the basis of discrete cosine transform, to construct a new transform mode with interval constraints to represent the high-speed rail seismic source seismic signals more sparsely, using the continuous wavelet transform for sparsely representing the strong background noise in the high-speed rail seismic source seismic data, and constructing an over-complete dictionary; separating and extracting the high-speed rail seismic source seismic signal and the strong background noise in the high-speed rail seismic source seismic data by using a block coordinate relaxation algorithm, and filtering the random white noise, so thatthe extracted high-speed rail seismic source seismic signal is more complete in structure and higher in signal-to-noise ratio.

The invention discloses a wind driven generator set failure feature extracting method based on sparse decomposition. The method is used for diagnosing wind driven generator set failures and is suitable for being used for non-direct-drive-type wind driven generator sets. The method is aimed to solve the problem that failure features of a wind driven generator set vibration signal are difficult to extract due to failure feature coupling, large background noise and the like, and on the basis of a form composition analysis principle, the signal is decomposed into three compositions including harmonic waves, impact and noise by means of different sparse expression dictionaries, so that the failure features which are difficult to find outstand; experiments are carried out on a wind driven generator failure simulation experiment table, and validness is proved. The method has the advantages that the failure features which are covered by noise or other feature signals and difficult to find are extracted, accordingly a failure portion is determined, instructive suggestions are made for maintenance of a wind driven generator set and arrangement adjustment of maintenance plans, non-scheduled down time is effectively reduced and economical efficiency and social efficiency of a wind power plant are improved.

The invention relates to a morphological component analysis (MCA)-based synthetic aperture radar (SAR) image noise suppression method. The method is technically characterized in that: an MCA method is applied to the suppression of multiplicative noises in an SAR image; and because the image generally comprises the two morphological components of edge / contour and texture, in the method, the dictionary assembly of curvelet and local discrete cosine transform (LDCT) is utilized under the uniform frame of MCA according to the good edge / contour representation characteristics of the curvelet and the high partial texture expression capability of the LDCT to fulfill the aim of suppressing the noises in the SAR image. Therefore, for a noise image, effective components of the image can be separated and the noises are kept in the remnant image by the method to fulfill the aim of suppressing the noises. Experimental results on both simulated and real SAR images show that the method has the advantage of relatively better noise suppression effect compared with the conventional wavelet-based methods and curvelet-based methods.

The invention discloses a geological radar B-scan data processing method. The method utilizes a signal processing method to process the B-scan data to separate and obtain data profiles for tunnel lining abnormality and railway roadbed disease detection. First, the continuum whitening spectrum expansion technology is used to improve the time resolution of the signal, and then the high-resolution data is migrated and imaged by using efficient phase shift migration to improve the lateral resolution of the section, so that the reinforcement inside the lining or roadbed on the imaging section Focusing on point structures, and finally based on the difference between point targets such as steel bars and reflection interfaces (concrete and clay layer or surrounding rock interface), the sparse separation method based on morphological component analysis is used to separate out different parts such as steel bar structures and reflection interfaces. morphological components. This method helps to improve the accuracy of tunnel lining and railway subgrade detection, and provides a strong basis for tunnel and railway maintenance.

The invention discloses a sparse optimization method of suppressing strong shielding interference of seismic data. The method includes: 01, reading original seismic data, and conducting a component analysis on the data; 02, determining two sparse representation dictionaries employed by a morphological component analysis according to the morphological characteristic difference between a seismic response of a transverse heterogeneous body including a river sand body etc. related to a reservoir and a seismic response of a stable sedimentary formation in a seismic record signal, and forming an ultra-complete dictionary; 03, extracting seismic response signals of the transverse heterogeneous body including the river sand body etc. related to the reservoir from a seismic profile after suppressing overlying or underlying formation strong reflection shielding by employing a block coordinate relaxation algorithm; and 04, repeating step 03 until completion of processing of all two-dimensional measuring line data. According to the sparse optimization method of suppressing the strong shielding interference of the seismic data, the strong shielding interference can be effectively and thoroughly suppressed, and the seismic response signals of the transverse heterogeneous body including the river sand body etc. related to the reservoir can be displayed more clearly and visually.