20 results about "Morphological filter" patented technology

The present application belongs to the technical field of image recognition, and particularly relates to a liquid crystal screen defect detection method based on image analysis. The method comprises: acquiring a liquid crystal screen surface image frequency domain signal and wavelet packet decomposition, and extracting a low frequency and medium-high frequency coefficient matrix; performing convolution on the low frequency coefficient matrix with an adaptive morphological filter kernel, fitting to generate a uniform illumination reference matrix; after weighted correction and denoising of the medium-high frequency coefficient matrix, inverse wavelet packet transformation is performed on the medium-high frequency coefficient matrix and the uniform illumination reference matrix to generate a reconstructed difference image; the local pixel gradient direction vector and the gray level co-occurrence matrix contrast feature of the reconstructed difference image are input into an isolation forest classifier to output a defect position and category. The present scheme shifts illumination separation to the frequency domain, avoids damage to defect pixels by spatial domain filtering, eliminates non-uniform illumination interference, retains the edge gradient and texture structure of small defects, and solves the problem of missing detection of small dark spots under a gradual illumination background.

The invention relates to a method for measuring the reactivity of pozzolanic materials and correlating this reactivity with strength development in composite cement. The method comprises dissolving pozzolanic materials in an alkaline solution and preparing a 1.0–5.0 M Sodium Hydroxide (NaOH) solution. This solution, containing the dissolved materials, is placed on a flow cell, configured to accommodate particles up to 100 µm. The flow cell is heated between 20 °C and 70 °C, and optical images of the particles are collected using an optical microscope. Particle count and area are measured by image processing with spatial and morphological filters to identify particles in the solution over time. The change in particle count or area is computed, and the normalized count represents dissolution plotted over time, describing the dissolution rate. This particle dissolution behaviour is correlated with compressive strength data obtained from samples tested within 1–28 days.

The invention relates to a variable-speed bearing fault feature extraction method and device, relates to the technical field of bearing fault diagnosis, and aims to solve the problem that the variable-speed bearing fault feature extraction precision is not high. The method comprises the steps that a particle swarm optimization algorithm is introduced into a multi-scale adaptive morphological filtering operator, key parameters of the adaptive morphological filtering operator are obtained, filtering signals of all scales are obtained through calculation according to the multi-scale adaptive morphological filtering operator and the key parameters, and the filtering signals of all the scales are added to obtain a primary filtering signal; introducing a particle swarm optimization algorithm into an adaptive stochastic resonance algorithm to obtain a system cooperative adjustment parameter of the adaptive stochastic resonance algorithm when the signal-to-noise ratio is maximum, and inputting the primary filtering signal into the adaptive stochastic resonance algorithm to obtain a secondary filtering signal; and calculating the secondary filtering signal by using an angular domain resampling algorithm to obtain an order spectrogram containing bearing fault features. The method has the effect of improving the fault feature extraction precision of the variable-speed bearing.

The invention relates to the technical field of image processing and computer vision, and discloses a cross-country road image segmentation and target integrated intelligent detection method comprising the following steps: S1, image input and gray level conversion; s2, carrying out Vibe background modeling and foreground extraction; s3, morphological filtering and denoising are carried out; s4, connected domain marking is carried out through an improved one-time scanning method; s5, optimizing a bounding box; s6, outputting a result; according to the method, the Vibe algorithm and morphological filtering are combined, dynamic noise such as vegetation shaking and light and shadow changes is effectively restrained, the accuracy and the anti-jamming capability of foreground segmentation are remarkably improved, the improved one-time scanning method is adopted for connected domain detection, the problem of target breakage is solved while the real-time processing efficiency is guaranteed, and the method is suitable for large-scale popularization and application. Through a bounding box optimization strategy, broken targets can be combined to ensure integrity, positions can be accurately restored, redundant boxes can be removed, and the method effectively aims at the complex environment of cross-country roads.

The present invention provides a voice recognition method, which includes the steps of: receiving a plurality of voice signals; performing a time-domain to frequency-domain conversion operation on the plurality of voice signals to generate a plurality of frequency-domain signals; using a morphological filter to perform filtering operations on the plurality of frequency-domain signals to generate a plurality of filtered backgrounds; generating a plurality of initial voice fingerprints according to the plurality of filtered backgrounds, wherein each of the plurality of initial voice fingerprints comprises times and frequency points corresponding to a plurality of peaks in the corresponding filtered background; generating at least one voice fingerprint according to the plurality of initial voice fingerprints; and storing the at least one voice fingerprint in a memory for subsequent voice recognition operation.

The application provides a ground point bottom model establishment method based on a point cloud minimum elevation clustering algorithm, which is divided into the following steps: open pit point cloud data collection, open pit point cloud data preprocessing, open pit point cloud data gridding, open pit point cloud bottom model establishment based on a DBSCAN density clustering algorithm, missing point cloud interpolation, and open pit final ground point bottom model point cloud data establishment. Specifically, first, open pit full point cloud data is collected, and the point cloud data is subjected to thinning processing; second, the point cloud data is gridded, all grid point cloud data is traversed, and the DBSCAN density clustering number of the lowest point in the grid is calculated in the order of elevation from low to high, and when the clustering number reaches N, the point is considered to be the minimum ground point; finally, missing point cloud is filled in through a nearest neighbor interpolation method, and then the final open pit ground bottom model point cloud data is obtained. The method aims to calculate the DBSCAN density clustering number of the lowest point in the grid, so that the open pit ground bottom model point cloud data can be quickly and accurately collected, which provides important guidance value for the establishment of an open pit DEM model by using an auxiliary progressive morphological filter and the intelligent acceptance of open pit blasting.

The invention belongs to the technical field of space flight and aviation remote sensing geological exploration, and particularly relates to a uranium ore altered mineral automatic extraction method for spaceborne hyperspectral data, which comprises the following steps of: 1, preprocessing the spaceborne hyperspectral data and optimizing wavebands; 2, constructing a multi-dimensional spectrum geometric feature set; 3, constructing a'net mineralization 'coupling model; step 4, carrying out adaptive threshold segmentation based on histogram curvature extreme values; and 5, morphological filtering post-processing is carried out, and a uranium ore altered mineral distribution diagram is output. According to the method, full-automatic and high-precision extraction of alteration information is achieved by constructing a positive and negative feature coupling model and combining histogram curvature extreme value analysis, and the problems that in existing hyperspectral mineral extraction, manual repeated threshold adjustment is seriously relied on, multi-index combined logic is rigid, and complex background interference is difficult to adapt are effectively solved.

The invention discloses a speech recognition system, which comprises a processing circuit and a memory. The processing circuit is configured to: receive a plurality of sound signals, respectively; respectively performing a time domain-to-frequency domain conversion operation on the plurality of sound signals to generate a plurality of frequency domain signals; using a morphological filter to filter the plurality of frequency domain signals, respectively, to generate a plurality of filtered background images; respectively generating a plurality of initial voice fingerprints according to the plurality of filtered background images, wherein each initial voice fingerprint in the plurality of initial voice fingerprints comprises time and frequency points corresponding to a plurality of peak values in the corresponding filtered background image; generating at least one voice fingerprint according to the plurality of initial voice fingerprints; and storing the at least one voice fingerprint in the memory for voice recognition.

The invention provides a voltage sag detection method and device based on a Teager energy operator and morphological filtering, and relates to the technical field of power system sag protection and electric energy quality monitoring. The method comprises the following steps: firstly, converting a three-phase voltage signal into a voltage component under a two-phase static coordinate system through Clark conversion; secondly, a virtual instantaneous voltage power modulus containing a fundamental wave offset mechanism is constructed, and sag characteristic components can be effectively extracted; then, a morphological filtering combination filter is used for conducting denoising processing on the differential power signals, high-frequency impulse noise interference is eliminated, and pure characteristic signals are obtained; and finally, calculating an energy spectrum of the pure characteristic signal by using a Teager energy operator, and carrying out voltage sag detection in combination with a self-adaptive dynamic threshold, so as to judge the starting and ending moments of the voltage sag. According to the invention, the voltage sag can be accurately detected within 1ms, and the anti-noise performance is excellent.

The application provides a morphological filtering image loss optimization method and device of an optoelectronic tracking system, and the method comprises the following steps: acquiring a basic morphological operation sequence formed by a morphological filtering algorithm adopted by the optoelectronic tracking system; determining a first pixel number required to be added and a second pixel number required to be deleted corresponding to each basic morphological operation based on the number of structure elements corresponding to each basic morphological operation; and performing optimization processing on each basic morphological operation, so that the optimized basic morphological operation performs the following operations: adding a corresponding first pixel number of pixels before and after a to-be-processed image sequence to obtain a reconstructed image sequence; performing the basic morphological operation on the reconstructed image sequence to obtain an operation result sequence; and deleting a corresponding second pixel number of pixels before and after the operation result sequence to obtain an output image sequence. According to the scheme, pixel loss can be effectively avoided, and the performance of the optoelectronic tracking system is improved.

The application provides a residual magnetism calculation verification method based on an arc starting point algorithm, comprising the following steps: collecting analog quantities of current and voltage and performing digital-analog conversion; when the collected data has an abnormal value, performing filtering processing on the sampling data by using a morphological filtering algorithm; detecting a closing period and judging whether a closing point exists, if not, returning to step S1, and if yes, executing step S4; extracting excitation inrush current waveform data and judging whether N continuous data are less than a current threshold value and the N+1th data is greater than the current threshold value, if not, returning to step S1, and if yes, outputting a mutation point and calculating a corresponding phase angle of the mutation point; and comprehensively calculating residual magnetism in a transformer core at the last time of switching off by derivation at the arc starting point. The method is simple and effective, is not affected by inaccurate magnetic flux estimation before and after magnetic saturation, requires only closing data, and can more accurately give a residual magnetism value.

The invention provides a morphological method and device for removing noise of a ship and a medium, and belongs to the technical field of seismic exploration. The method comprises the following steps: step 100, inputting seismic shot gather data; step 200, determining a parallel minimum granularity unit by adopting an SPARK parallel framework, and forming a target set of morphological filtering calculation; step 300, constructing structural elements of morphological filtering, and then predicting noise; step 400, performing boundary noise process treatment; and step 500, separating interference noise of the large ship to obtain de-noised seismic shot gather data. According to the method, two-dimensional mathematical morphology filtering and boundary processing of bidirectional structure elements are combined for denoising calculation, interference noise can be accurately predicted, denoising is accurate, parallel calculation is adopted, the calculation amount can be greatly reduced, the calculation efficiency is greatly improved, and the calculation cost is saved.

This invention relates to the field of signal source monitoring, specifically to a train operation status monitoring method based on DAS spatiotemporal sequence imaging. The invention employs a time-domain frame-segmented wavelet packet feature extraction method to preprocess the train vibration signal, removing trailing vibration signal interference. Then, based on image binarization and morphological filtering methods in image processing algorithms, the train trajectory is feature-enhanced. Image segmentation is used to detect local straight lines, k-means clustering is used to classify the corresponding straight lines of the train, and polynomial fitting is used to extract the train trajectory. Finally, based on the obtained train trajectory, the wavelet packet energy feature value matrix is ​​truncated, and an adaptive threshold method is used to obtain the train's forward and backward trajectories, thereby achieving accurate monitoring of train length and speed, effectively solving the accuracy problem of existing DAS systems in monitoring train operation status.

The invention provides a signal de-noising processing method, device and equipment and a storage medium, and the method comprises the steps: carrying out the wavelet de-noising of an electric signal based on a wavelet transform filter, and obtaining an electric signal after the wavelet de-noising; and weakening a pseudo-Gibbs effect in the electric signal after wavelet denoising based on a morphological filter to obtain a target electric signal without myoelectricity noise. According to the method and the device, effective processing and purification of the electric signals are realized through measures such as wavelet denoising based on wavelet transformation and pseudo-Gibbs effect weakening by the form filter, particularly, myoelectricity noise possibly existing in the electric signals is further removed through the form filter, the purity and the quality of the target electric signals are improved, and the method and the device are suitable for popularization and application. And powerful support is provided for subsequent analysis and application of the target electric signal.

This invention belongs to the technical field of wear condition identification for hot forging press components, specifically relating to a feature signal enhancement method for wear condition identification of hot forging press components, including: S1, enhancing the feature signal x of inferior vibration signal x poor The repair signal x is obtained through processing. re S2. Construct a closed-open self-complementary top-hat morphological filter FCO-NSTH to enhance the characteristic frequency amplitude of the vibration signal to be enhanced; S3. Adaptively select the optimal structural element parameters of the closed-open self-complementary top-hat morphological filter using a particle swarm optimization algorithm; S4. Use the closed-open self-complementary top-hat morphological filter FCO-NSTH with adaptively selected optimal structural element parameters to enhance the repair signal x. re The characteristic frequency amplitude is enhanced to obtain an enhanced vibration signal; S5. The enhanced vibration signal obtained in S4 is input into the intelligent model to identify and predict the wear state of hot forging press components. This method can accurately identify the wear state of hot forging press components by combining with the intelligent model.

The application provides a method and related equipment for enhancing a while-drilling seismic signal. The method comprises: acquiring a while-drilling seismic signal; calculating a time-varying structure element according to a local variance of each trace of the while-drilling seismic signal; performing morphological dilation and corrosion operations at different scales for each time point and spatial trace of each while-drilling seismic signal to calculate a multi-scale morphological gradient value; constructing a weight matrix according to the multi-scale morphological gradient value; and performing weighted morphological filter iteration reconstruction on the while-drilling seismic signal based on the weight matrix to obtain an enhanced while-drilling seismic signal. The embodiment of the application realizes real-time enhancement of the while-drilling seismic signal by dynamic structure element generation and multi-scale morphological gradient weighting, and combines weighted morphological filter iteration reconstruction, thereby significantly improving the signal-to-noise ratio, retaining lithology interface reflection characteristics, suppressing noise and interference, and providing reliable technical support for underground coal mine seismic monitoring.

The application discloses a kind of fusion multi-scale morphological and adaptive energy threshold electrocardiosignal processing method and system, comprising: obtaining electrocardiosignal, utilize multi-scale morphological filter, to electrocardiosignal Parallel signal path calculation and baseline path calculation, the morphological reconstruction signal of weighted subtraction calculation result is obtained;By non-linear energy operator to morphological reconstruction signal is amplified, and energy sequence is obtained;Energy value in energy sequence is detected gradually, if the energy value is greater than main threshold, then directly confirm as R wave, and update main threshold;If the energy value is less than main threshold, judge distance last R wave time whether greater than 1.5 times average RR interval, if not greater than, then read next energy value;Otherwise, by backtracking search to obtain backtracking maximum value, judge backtracking maximum value whether greater than auxiliary threshold, if greater than, then mark as supplementary R wave;Otherwise, read next energy value;Based on confirmed R wave, output denoised electrocardiosignal waveform and heart rate parameter.

The invention relates to the technical field of electrical variable measurement, and discloses a charging pile output current characteristic on-line extraction method comprising the following steps: synchronously obtaining a current sampling sequence and a control duty ratio sequence; determining a current rising slope and a current falling slope of the energy storage inductor according to the duty ratio and the power conversion parameter; determining a left side horizontal span and a right side horizontal span of a filtering operator according to a ratio of the rising slope to the falling slope, and constructing an asymmetric morphological filtering operator; the asymmetric morphological filtering operator is used for processing a current sampling sequence, extracting a non-periodic transient characteristic edge and packaging, power control parameters are used for reconstructing the geometrical morphology of the filtering operator, physical matching of the measurement operator and sawtooth ripple topology is achieved, waveform inflection point distortion caused by a symmetric window is eliminated, and the measurement accuracy is improved. The millisecond transient characteristic envelope integrity is maintained while the switch ripple is suppressed, and high-fidelity electric heating boundary data is provided for loop health assessment.

This invention discloses a method for outlier detection in multi-beam data based on adaptive progressive morphological filtering. The method employs a closed-open-open combined average morphological operator for filtering, extracting a local terrain slope tolerance parameter. Based on this parameter, the elevation threshold is adaptively and dynamically adjusted during iterative morphological filtering. Through multi-level filtering windows and iterative calculations, the morphological reference plane gradually approximates the actual seabed until the iteration termination condition is met. Then, the terrain slope is extracted and a global judgment threshold is calculated. The depth deviation between the actual measured depth and the reference elevation is compared based on spatial location, thereby achieving accurate removal of outliers from the original discrete point cloud. By establishing a local dynamic adjustment and global decision mechanism for the filtering threshold, the method effectively avoids the filtering and under-filtering defects caused by fixed parameters, improving denoising accuracy while ensuring high fidelity of the actual seabed micro-topography.