74 results about "Mathematical morphology" patented technology

The invention discloses an underwater target image recognition method based on feature selection. The method mainly comprises six steps. Firstly, an original image of an underwater target is obtained, and median filtering is applied for denoising to enhance the edge of the image; then, contrast enhancement and sharpening processing are carried out on the denoised image, and target edge features are optimized in combination with mathematical morphology operation; secondly, extracting color features, texture features and edge features of the image, and performing normalization processing on the color features, the texture features and the edge features; the features are selected and dimensionally reduced through a kernel principal component analysis (KPCA) method. And based on the dimension-reduced features, constructing an artificial neural network for classification by using a multi-layer perceptron (MLP), and optimizing the model through a back propagation algorithm. And finally, the classification result and the performance index of the model are verified through experiments to ensure the effectiveness of the method. According to the method, automatic identification of the underwater target can be realized, and effective technical support is provided for related fields.

The invention discloses a lung three-dimensional reconstruction method and system for intraoperative positioning, and the method comprises the steps: processing a lung CT image in a DICOM format through an improved lung parenchyma segmentation algorithm, precisely extracting a lung structure through a global threshold method, connected region analysis and mathematical morphology processing, and then processing a mask generated through segmentation through a real-time smooth interpolation method, and through a dynamic parameter adaptation technology, accurate matching of a segmentation result and a gray value range of modeling data is ensured. According to the method, the problems of artifacts, redundant surfaces and gray level jump of the lung parenchyma area caused by separation of segmentation and reconstruction in a traditional method are avoided, full-process automation from data reading and lung parenchyma segmentation to three-dimensional modeling and visualization is achieved, meanwhile, a user-friendly interaction interface is designed, and a convenient tool is provided for intraoperative navigation.

The invention discloses a multispectral remote sensing accurate extraction method for an offshore wind turbine point location, and the method comprises the steps: 1, collecting a multi-temporal satellite remote sensing image covering a coastal zone region, carrying out the water body information enhancement and threshold segmentation processing, generating a sea-land binary image, and separating an offshore research region; 2, collecting and dividing a multi-stage remote sensing data set of the offshore wind plant; constructing an offshore wind turbine enhancement index capable of effectively distinguishing the offshore wind turbine and the background ground feature, and generating an offshore wind turbine enhancement image; constructing a visible light superposition index capable of retaining shadow information of the offshore wind turbine, and generating an offshore wind turbine shadow enhanced image; 3, multi-directional gray-level co-occurrence matrixes are calculated for the multi-stage offshore wind turbine enhanced images respectively, and a low-noise wind turbine binary image is obtained through synthesis; and 4, performing mathematical morphology expansion processing and vectorization on the low-noise fan binary image, and performing calculation to obtain accurate position data of the offshore fan.

The present invention belongs to the technical field of machine vision detection, and relates to a method for detecting sizing yarn hitting a reed based on machine vision. The present invention provides a method for detecting sizing yarn hitting a reed based on machine vision, which can detect and track the phenomenon of warp yarn hitting a reed during the sizing process, and is a method for detecting sizing yarn hitting a reed with high accuracy, strong robustness, and wide variety adaptability. The complete image of the warp yarn passing through the reed teeth is captured by an image acquisition system that is independently built and can adapt to different types of sizing machines. First, the initial image is cropped to obtain the detection area, and then the image is smoothed by Gaussian filtering; finally, by improving the inter-frame difference method, expanding the inter-frame difference, adding the Canny edge algorithm and mathematical morphology operation, the detection and tracking of the warp yarn hitting the reed are realized; at the same time, the proposed algorithm can be applied to detection and recognition problems in other fields.

The invention relates to a city expansion mode identification method fusing mathematical morphology and landscape connectivity, and belongs to the crossing field of city planning and geographic information technology. The method comprises the following steps: identifying corridor type expansion based on road data; urban holes are recognized through seed filling and morphological closed operation, and then filling type expansion located in the holes is recognized; calculating the minimum expansion step number based on eight-neighborhood iterative expansion to recognize enclave type expansion; and the connectivity and the linear form are fused to identify gallery type expansion and edge type expansion. Compared with the traditional landscape expansion index LEI, the classification precision of the morphological landscape expansion index MLEI provided by the invention is remarkably improved, and the method is suitable for multi-scale urban evolution analysis.

The invention discloses a hydraulic pump fault detection method and device based on information fusion and a medium, and belongs to the technical field of hydraulic device fault detection. The method comprises the steps of collecting a vibration signal and a pressure signal of the hydraulic pump; processing the vibration signal and the pressure signal based on a mathematical morphology filtering algorithm to generate a de-noised vibration signal and a de-noised pressure signal; processing the signals through a convolutional neural network to generate a vibration feature map and a pressure feature map; generating a weighted vibration feature map and a weighted pressure feature map based on an attention mechanism; integrating the weighted vibration feature map and the weighted pressure feature map, and generating a comprehensive fault feature vector; generating normalized comprehensive features based on a normalization algorithm; and processing the normalized comprehensive features through a full connection layer and a Softmax classification layer of the convolutional neural network, and outputting a fault category diagnosis result. By means of the method, the technical effect of improving the accuracy and reliability of fault detection of the hydraulic pump is achieved.

The invention discloses a mathematical morphology-based isolation switch icing thickness image identification method, relates to an isolation switch icing thickness image identification method, and solves the problem of poor precision when an existing edge detection method is used for identifying an isolation switch icing thickness image. The method comprises the following steps: acquiring an ice-coated isolation switch image; carrying out image preprocessing on the ice-coated isolating switch image to generate a gray-scale enhanced image; performing edge extraction on the grayscale enhanced image by using mathematical morphology to obtain an edge image; performing edge selection on the edge image to obtain an ice-coated isolation switch extraction image; and comparing and calculating the obtained original edge data of the uniced isolation switch with the extracted image of the iced isolation switch, and calculating the icing thickness of the isolation switch. The beneficial effect is that the icing thickness of the isolation switch can be accurately obtained.

The invention relates to the technical field of municipal pipe networks, in particular to a water supply pipe network zoning metering area dividing method and device and electronic equipment. The method includes: acquiring digital elevation model data of a target area; performing depression filling on the digital elevation model data by adopting a mathematical morphology method to generate a depression-free elevation model; performing watershed division on the depression-free elevation model based on a hydrological analysis algorithm to obtain river network distribution characteristics and sub-basin division results; combining the river network distribution characteristics and the sub-basin boundaries to generate a preliminary partition metering region; performing spatial overlay analysis on the sub-basin division result and geographic information data to obtain water supply network management parameters; and adjusting the preliminary zoning metering area according to the water supply network management parameters, and outputting a final zoning result. The reliability of the division result is high, and support is provided for construction of urban water supply networks.

The invention discloses a forest airborne laser radar point cloud cutting method, device and equipment, a medium and a product, and relates to the field of forest airborne laser radar point cloud data processing.The method comprises the steps that according to a tree crown height model, tree crown edge features are quantized through mathematical morphology opening operation; the crown height model is constructed based on height information in forest airborne laser radar point cloud data; distance features are quantized according to the horizontal distance between the pixels in the crown height model and the neighborhood pixels of the crown height model; fusing the crown edge features and the distance features, and determining fusion features; constructing a connected graph according to the fusion features, and taking the shortest path between end points in the connected graph as a cutting line; and according to the cutting line, performing block processing on the crown height model, determining a plurality of block crown height models, and extracting forest airborne laser radar point cloud data in the block crown height models, thereby improving the cutting precision of the laser radar point cloud data of the crown in the forest scene.

The invention discloses a variable-scale multispectral remote sensing image segmentation method using a spectral index, which belongs to the technical field of remote sensing image processing, and comprises the following steps: firstly, determining the gradient of a multispectral remote sensing image by using a vector field model, generating a mark according to the gradient, selecting a proper spectral index, and carrying out image segmentation according to the spectral index; and carrying out image segmentation by using watershed transformation. According to the variable-scale multispectral remote sensing image segmentation method using the spectral index, optimization is carried out by using the spectral index, the capability of distinguishing different ground feature types is enhanced, and the segmentation precision is improved; according to the method, watershed transformation is adopted for segmentation, so that the image segmentation result is more stable and efficient, the method can adapt to remote sensing images of different scales, the quality and effect of remote sensing data processing are improved, and mathematical morphology-based variational scale watershed segmentation is optimized.

The invention discloses an offshore wind power direct current collection line fault positioning method based on traveling wave multi-scale gradient analysis. The method comprises the steps that S1, a protection element judges that a fault occurs; s2, sampling the voltage and current of the positive electrode and the negative electrode of the bus side of the fault collection line, and calculating a 1-mode voltage reverse traveling wave; s3, filtering the 1-mode voltage reverse traveling wave based on a mathematical morphology algorithm, and calculating a filtering signal; s4, calculating a multi-scale gradient of a filtering signal, identifying a transient state abrupt change point of a waveform, and extracting arrival time and wave head polarity information of a fault voltage traveling wave; and S5, calculating a fault distance. According to the method, fault positioning is realized by using the fault voltage data within a short time after the fault, the specific fault position of the collection line can be quickly determined, and the method is of great significance in accelerating fault clearing and guaranteeing safe and stable operation of an offshore wind power system.

The invention particularly relates to an AI auxiliary diagnosis method for oral cavity observation, which relates to the technical field of oral cavity medical AI and medical image security, and comprises the following steps: generating a Hash watermark by using patient information, and embedding the watermark into an intermediate frequency alternating current coefficient of a diagnosis core area by adopting a coefficient relation modulation method. According to the method, the local terminal is subjected to deterministic partitioning and frequency domain irreversible desensitization processing, rigid protection of biological privacy of a patient is achieved from the source, and the risk of privacy leakage in oral medical image transmission and storage is dealt with; color space conversion, threshold screening and mathematical morphology operation are completed at a local terminal, a diagnosis core area and a privacy sensitive area are accurately divided, and privacy processing can be completed without uploading an original image; the privacy sensitive area is subjected to combined processing of direct current component retention, medium and low frequency coefficient pseudo-random attenuation replacement and high frequency coefficient zero setting, a reverse mapping table and attenuation factors are not retained, irreversible blinding is achieved on the mathematical level, and biological characteristics such as cheilogramma and facial skin cannot be restored.

The application discloses a kind of offshore wind power dc collection line fault location methods based on travelling wave multiscale gradient analysis, S1.distinguishes that fault occurs;S2.sampling the voltage and current of positive pole and negative pole of fault collection line bus side, calculate 1 mode voltage reverse travelling wave;S3.based on mathematical morphology algorithm, 1 mode voltage reverse travelling wave is filtered, and filter signal is calculated;S4.the multiscale gradient of filter signal is calculated, and the transient abrupt point of waveform is identified, and the arrival time and wave head polarity information of fault voltage travelling wave are extracted;S5.calculating fault distance.The method of the application realizes fault location using fault voltage data in a short time after fault, and can quickly determine the specific fault position of collection line, and has important significance for accelerating fault clearing and guaranteeing the safe and stable operation of offshore wind power system.

The application discloses a kind of based on perception voxelization three-dimensional model generation and solidification method, it is related to artificial intelligence three-dimensional model generation technical field, comprising: S1, multi-modal three-dimensional asset input and analysis;S2, three-dimensional model surface sampling and color extraction;S3, three-dimensional space normalization and scale constraint;S4, three-linear weight sputtering based on voxel gridding;S5, voxel structure repair based on mathematical morphology;S6, internal cavity identification based on Flood Fill;S7, conservative solidification processing based on distance transformation;S8, color quantization based on perception color space;S9, sparse voxel data generation and model output.The application, through unified analysis interface, improves the compatibility and stability of multi-source three-dimensional asset conversion;Traditional attribution is replaced by three-linear weight sputtering, and sampling points are distributed to eight-neighborhood voxels according to distance, which reduces aliasing and ensures continuous expression of slender structures at low resolution.

The invention provides a feature extraction method and device of a thyroid cancer image based on a convolutional network, electronic equipment and a storage medium, and the method comprises the steps: obtaining a preset number of color Doppler ultrasound images of thyroid cancer, and carrying out the preprocessing of the color Doppler ultrasound images; extracting a thyroid cancer image from the preprocessed color Doppler ultrasound image, performing color space conversion on the thyroid cancer image through a preset color space, and segmenting a blood flow spectrum image in the thyroid cancer according to hue and saturation information in the color space; performing a mathematical morphology operation on the blood flow spectrum image to enhance the blood flow spectrum image; and inputting the blood flow spectrum image after the mathematical morphology operation into the trained feature extraction model for convolution processing so as to obtain the proportion and distribution condition of the blood flow spectrum signal in the thyroid cancer. Therefore, effective features in the thyroid cancer image can be accurately identified, and thyroid cancer and thyroid benign nodules can be effectively distinguished.

The present invention discloses an unsupervised color geological image segmentation method and system based on an attention mechanism. First, the color geological image is preprocessed by mathematical morphology to obtain a preprocessed image. Then, the preprocessed image is respectively processed by the SE-UNet neural network and the Felzenszwalb clustering algorithm to obtain a rough segmentation image and a superpixel segmentation map. Finally, the superpixel segmentation map is used as a "pseudo-label" to optimize the rough segmentation image based on the superpixel segmentation map, and the predicted segmentation result is output through forward propagation, realizing the segmentation of unlabeled color geological images. The present invention segments color geological images without any manual annotation and has strong automatic segmentation ability.

The invention discloses a cable joint early fault detection method based on mathematical morphology, and belongs to the technical field of power line control, and the method comprises the steps: introducing a carbonization channel resistor and a water film resistor on a diode model, and obtaining a fault equivalent circuit model; building a cable network structure in the simulation software, setting a single-phase earth fault of the damped cable joint in the cable network structure, and performing fault simulation by using the fault equivalent circuit model; based on a fault simulation result, determining a zero-sequence voltage as an early fault detection signal, and enhancing and capturing fine distortion in a zero-sequence voltage waveform by using top-hat transformation to obtain a top-hat characteristic quantity; if the top cap characteristic quantity is greater than an amplitude safety threshold value and the duration greater than the amplitude safety threshold value is greater than a safety duration threshold value, judging that an early fault occurs; and determining the development stage of the early fault by using the top hat feature relative variance. According to the method, the early fault of the damped cable joint is accurately identified, the development stage of the damped cable joint is judged, and the detection sensitivity is improved.

This invention discloses a method for extracting moiré fringe phase information based on Morlet wavelet transform, comprising: converting a target moiré fringe image into a grayscale image; filtering the grayscale image to obtain a filtered binarized image; performing mathematical morphological denoising on the binarized image to obtain a processed image, wherein the mathematical morphological denoising includes erosion and dilation; convolving each column of the processed image using modulated Morlet wavelets and calculating the amplitude and phase of the wavelet coefficients; determining wavelet ridges based on the amplitude of the wavelet coefficients and extracting the phase of the scale coefficients and time coefficients corresponding to the wavelet ridge positions; and obtaining complete moiré fringe phase information based on the phases of all extracted columns. This method is highly accurate and easy to implement, making it suitable for further refractive index reconstruction and key parameter measurement of the measured flow field.

The invention provides an active power distribution network fault section positioning method based on mathematical morphology polarity, and relates to the technical field of power distribution network fault positioning. The method comprises the following steps: when any phase of measurement current in the feeder terminal meets a fault detection criterion, considering that a fault may occur, entering a fault positioning process based on mathematical morphology polarity, and executing the following steps: the feeder terminal selects a proper morphological M value according to the number of fault phases; filtering the measured current waveform by using a mathematical morphology algorithm; after the filtering is completed, calculating a fault current break variable by using a full-cycle subtraction method; extracting a peak value of the fault current break variable waveform by using a mathematical morphology algorithm, and judging the polarity of the peak value; terminal devices at the two ends of the protected section mutually transmit the polarity of the mathematical morphology peak value; and considering the section as a fault section according to the fact that the polarities of mathematical morphology peak values at two ends are regular. The method is not influenced by the fault characteristics of the distributed power supply, and can better adapt to the complex fault characteristics of the active power distribution network.

The invention discloses an automatic cell analysis and counting method based on mathematical morphology, and relates to the field of cell analysis and counting. The method comprises the following steps: acquiring an image; preprocessing the image; expansion, hole filling, corrosion treatment and filtering; carrying out fracture treatment; performing a connected domain algorithm to obtain a cell connected domain; preliminary screening; counting according to the required cell area, cell shape and cell state, and finally obtaining the number of cells in the image. The method not only improves the accuracy of cell counting, but also is simple and convenient to operate, and greatly improves the efficiency and reliability of cell analysis.

The present invention belongs to the field of optical engineering, and specifically relates to a method for adjusting the optical power symmetry of a cone assembly of a double-probe illuminated laser module, including the following steps: (1) designing a device for adjusting the symmetry of the laser module cone assembly; (2) collecting characteristic images at the bottom of the laser module; (3) designing a lighting scheme; (4) image processing; (5) target feature extraction; (6) image angle compensation; (7) adjusting the optical power symmetry; The present invention uses the method of energizing the double probes, completely energizing from the bottom of the laser module, so that the laser module with an insulated outer cylinder can be normally illuminated in the improved double-probe JAS-15 chuck; Four groups of optical power acquisition components clamp the directions of two strong lights and two weak lights, which can effectively improve the adjustment accuracy and shorten the adjustment time; The sobel operator is used to extract the target contour, and the mathematical morphology method is used to process the target contour, and the edge integrity of the extracted graphic target features is better.

The present invention relates to the field of visual detection technology, and in particular to a chip visual positioning system and method based on image contour feature and grayscale matching, wherein an image acquisition module is moved to a chip detection position to acquire an image of a chip to be tested, and an image of the chip to be tested is obtained; an image preprocessing module is used to preprocess the image of the chip to be tested based on mathematical morphological operations, so that the outline of the chip gate bonding area is clear; an image matching positioning module is used to extract the outline of the chip gate bonding area of ​​the preprocessed chip image to be tested, and the center point coordinates and rotation angle of the chip gate bonding area are calculated. Based on the center point coordinates and rotation angle, a local rectangular search area and a matching angle are designed as a search window, and an improved NCC visual positioning method is used to accurately locate the center point of the chip gate bonding area, and a visual positioning detection result is obtained. The present invention can improve the speed, accuracy and robustness of chip visual positioning.

This invention relates to a method and apparatus for bright-field cell image segmentation in fluorescence microscopy. Combining an improved two-dimensional OTSU threshold segmentation algorithm, it filters out noise points in single-cell images during image preprocessing and further refines cell segmentation using binary image mathematical morphology. Then, it segments adhered cells using a label-controlled watershed segmentation algorithm based on cell nucleus images. This invention improves segmentation results through image enhancement and refines them using various segmentation methods, effectively addressing issues such as weak edges, poor contrast, irregular cell shapes, and cell adhesion in bright-field cell images. Therefore, the overall visual segmentation effect is quite good.

The present invention relates to an electrocardiogram key band extraction method, device and medium. The electrocardiogram preprocessing method comprises: obtaining an electrocardiogram to be processed, determining an electrocardiogram signal to be processed according to the electrocardiogram to be processed; performing baseline drift removal processing on the electrocardiogram signal to be processed based on a mathematical morphology filtering algorithm to obtain an electrocardiogram signal without baseline drift; performing power frequency interference removal processing on the electrocardiogram signal without baseline drift based on a Wiener filtering algorithm to obtain an electrocardiogram signal without power frequency interference; performing myoelectric interference removal processing on the electrocardiogram signal without power frequency interference based on an empirical mode analysis method and a principal component analysis to obtain an electrocardiogram signal without myoelectric interference; performing motion artifact removal processing on the electrocardiogram signal without myoelectric interference based on an empirical mode analysis method and a Welch power spectrum analysis method to obtain an electrocardiogram signal without motion artifact. The present invention is used to remove noise from an electrocardiogram and improve the accuracy of electrocardiogram diagnosis.

The invention discloses a metallized film capacitor self-healing area automatic identification method, device and system and application thereof. The method comprises the following steps: acquiring image information of a capacitor metallized film; performing graying processing on the image information to obtain a grayed image; and carrying out image binaryzation, image denoising and contour detection on the grayed image to obtain parameter information of a self-healing area of the capacitor metallized film. According to the method, the self-healing area is accurately recognized based on image recognition of mathematical morphology, and the reliability and consistency of detection are improved.

The application discloses a detection method, a storage medium and a device based on an OPENCV reagent cover opening angle, and the method comprises the following steps: (1) taking a picture of a hole plate area before each injection to obtain an image A; (2) pre-processing the image to obtain an effective range image B; (3) performing gray scale processing on the image by using the OPENCV to obtain a graph C; (4) performing binary processing on the image by using the OPENCV to obtain an image D; (5) performing mathematical morphology processing on the image to obtain an image E; (6) designing an elliptical detection area to obtain an image G; (7) selecting a detection area; (8) judging a reagent cover state; the application realizes automatic judgment on the opening and closing state of a reagent Tip head cover, calculates the opening and closing angle of the cover and makes angle marking, eliminates the cost of manual judgment, improves the reagent injection efficiency, and makes the stability and consistency of the whole production process higher.

The invention discloses a constant-resistance anchor cable supporting effect on-line monitoring method and system, particularly relates to the technical field of deformation state on-line monitoring, and is used for solving the problems that an existing method based on single-point monitoring cannot analyze multi-anchor cable collaborative response from a space-time correlation level, and a rock stratum activity energy release core area is difficult to position. The method comprises the following steps: acquiring and registering working data of an anchor cable network and geological structure information, implementing high-precision time synchronization and delay compensation on the working data of the anchor cable to form a networked time sequence data set, and identifying a time-space intensive response event cluster from the networked time sequence data set; coupling degree analysis is carried out on the response event cluster and the geological structure to confirm a multi-anchor-cable cooperative response event, then a space-time two-dimensional data field is constructed based on the space-time range of the event, multi-scale mathematical morphology gradient calculation is carried out on the data field online, and the multi-anchor-cable cooperative response event is obtained by tracking the space-time convergence area of the local maximum value of each scale gradient field. And finally, determining an energy release focus of the multi-anchor-cable collaborative response event.

This application discloses a bearing fault diagnosis method, equipment, medium, and product, relating to the field of fault diagnosis. The method includes: acquiring bearing information data; performing mode decomposition on the information data based on mathematical morphology and using an empirical wavelet transform spectral partitioning method to obtain a set of mode components; calculating the autocorrelation kurtosis of any mode component in the set; determining an autocorrelation kurtosis feature vector based on all autocorrelation kurtosis values; inputting the autocorrelation kurtosis feature vector into a fault diagnosis model and outputting a diagnosis result; the diagnosis result is either normal or a fault exists; the fault diagnosis model is trained based on a convolutional neural network. This application enables feature extraction and intelligent diagnosis of bearing faults.

The invention relates to the technical field of image recognition, and provides an intelligent food detection method and system based on image recognition, and the method comprises the steps: collecting a corn kernel image, and constructing a distance transformation diagram; all local maximum value points in the distance transformation diagram are screened out, a centroid height threshold value and a concave point height threshold value are set, mathematical morphology processing is carried out on the distance transformation diagram according to the centroid height threshold value and the concave point height threshold value, and centroid pixel points and concave points are obtained; determining adjacent centroids of the pits, and establishing a rigid mark graph according to the remaining pits and the positions of the adjacent centroids of the pits; image segmentation is carried out according to the rigid mark graph and the distance transformation graph, a segmentation label graph is obtained, areas corresponding to the corn kernels in the segmentation label graph are extracted, detection of the corn kernels is completed according to contour features of the areas corresponding to the corn kernels, and the corn kernels are divided into different quality types. According to the invention, the accuracy of corn kernel identification can be improved.

The application belongs to the technical field of fault line selection, and particularly relates to a small current secondary grounding fault line selection method and system based on mathematical morphology. Bus zero sequence voltage and each feeder zero sequence current are collected. The voltage signal is decomposed by using a mathematical morphology OCCO algorithm to extract transient and steady components. If the transient component exceeds a threshold value, voltage disturbance is determined and the fault time is recorded. If the effective value of the steady component exceeds the threshold value after the fault time, it is confirmed that a single-phase grounding fault occurs. According to the zero sequence current of each feeder after the voltage disturbance occurs, the transient component is obtained by using the OCCO algorithm, and the transient energy value of each feeder is calculated accordingly. The feeder with the maximum energy is selected as the suspected fault line. The instantaneous transient power of the transient component of the current and the transient component of the voltage of the line is calculated, and the fault type is determined according to the polarity. The OCCO algorithm adaptively extracts fault features, and the double criteria of transient energy and instantaneous power polarity can accurately identify single-phase grounding and secondary grounding fault lines.