28 results about "Directional filtering" patented technology

The invention discloses a surface water reserve monitoring method and system based on satellite gravity data, electronic equipment and a storage medium, and the monitoring method comprises the steps: S1, obtaining the satellite gravity data, and carrying out the omni-directional filtering and denoising of the satellite gravity data, and obtaining the preprocessed gravity data; s2, performing time domain and frequency domain decomposition on the preprocessed gravity data, and extracting surface water reserve change signals on different spatial and temporal scales based on time domain and frequency domain information; and S3, constructing a surface water level virtual change value based on the surface water reserve change signals on different spatial and temporal scales, and performing analog calculation based on the surface water level virtual change value to complete surface water reserve monitoring.

The invention relates to the technical field of earphone noise reduction, and discloses an active noise reduction method and device for a head-mounted wireless earphone and the head-mounted wireless earphone. The method comprises the following steps of: performing time division multiplexing sampling through double microphone arrays of left and right acoustic cavities of the earphone to obtain four paths of microphone array sampling data and performing correlation function beam forming calculation to generate sound source direction distribution data; calculating a geometric acoustic path difference of the earphone, and establishing a direction delay parameter table corresponding to each effective sound source direction; driving a multi-channel adaptive filter bank according to the direction delay parameter table, performing directional filtering, and generating a directional inversion signal group; and taking the intensity information in the sound source direction distribution data as a weighting coefficient to perform weighted synthesis on the directional inversion signal group, forming a comprehensive inversion signal, applying a feedforward delay correction value, and outputting a self-adaptive noise reduction signal. According to the method, continuous and stable noise reduction performance can be kept in a complex dynamic multi-sound-source environment, and the auditory experience of a user is remarkably improved.

This invention discloses a system and method for labeling automotive abnormal noise data, relating to the field of automotive NVH detection technology. The method includes: receiving raw signal data of automotive abnormal noise; performing time-frequency transformation on the raw signal to generate a time-frequency diagram; identifying key frequency ranges characterizing the abnormal noise features; then performing directional filtering to enhance the abnormal noise signal; identifying candidate time ranges for abnormal noise through an automatic detection algorithm; and outputting the final abnormal noise time range labels and key frequency ranges in a structured format after verification and correction. The system includes a data receiving module, a signal processing module, a display module, an intelligent labeling module, an audio playback module, a labeling output module, and an interaction module, achieving triple-assisted labeling through auditory, visual, and intelligent pre-selection methods. This solves the problems of low efficiency, low accuracy, and poor consistency in traditional manual listening methods, providing high-quality labeled data for deep learning classification of automotive abnormal noises.

The present invention discloses a method and system for enhancing image quality, relating to the technical field of image processing. The method includes: extracting the luminance component of each frame from a video stream, calculating the resultant image after directional filtering of the luminance component through boundary filtering; calculating the image after non-directional filtering of the luminance component; calculating the overall gradient range of the image, and integrating the resultant image after directional filtering and the image after non-directional texture filtering through the overall gradient range. The method and system disclosed by the present invention enable the retention of image boundaries and the thorough removal of noise.

The present invention relates to a video flow measurement method and device based on LK tracking and grayscale statistical feature methods, including using the LK tracking algorithm to track speed measurement points to form speed measurement point tracking trajectories for the first ten frames of a video; enhancing the texture of the spatiotemporal image trajectory and suppressing background noise and stripe direction noise using a Gaussian directional stretching filter, a noise suppression function, and a directional filtering function; using a random Hough transform to detect the straight angles of the spatiotemporal image to obtain an approximate rotation range; statistically calculating the grayscale histogram of each angle of the rotated spatiotemporal image; calculating the eigenvalue of the grayscale histogram; and drawing a characteristic curve to obtain the optimal texture angle. The present invention can automatically mark speed measurement lines, effectively remove noise from the spatiotemporal image and enhance the trajectory, identify clear spatiotemporal image texture, and thereby improve the robustness, applicability, and accuracy of the spatiotemporal image measurement method, effectively overcome the influence of the complex environment in natural river channels on the measurement results, and improve detection accuracy.

This invention discloses a deep learning method for detecting water leakage defects based on feature input enhancement. The method includes: directional filtering and suppression of water leakage images, using the directional filtering results as the first part of the image feature channel input to a subsequent deep neural network; segmentation and enhancement of water leakage region signals: tunnel inspection image data is segmented and preprocessed using the Otsu threshold segmentation method, followed by smoothing using an image spatial domain filter; multiplication of the feature input enhancement weight matrix after spatial domain filtering with the original image, and using the resulting result as the second part of the image feature channel input to a subsequent deep neural network; and establishment of a water leakage detection and recognition model. This method significantly improves the performance of water leakage detection and recognition.

The invention discloses a window optimization angle filtering anomalous body trend determination method and system, and the method comprises the steps: building an anomaly model composed of at least three anomalous bodies, and adding a preset noise amount of Gaussian white noise into the anomaly model, so as to obtain magnetic anomaly data; and performing model verification on the magnetic anomaly data by adopting multiple filtering algorithms to obtain an optimal filtering algorithm, and performing filtering processing on the target magnetic anomaly data based on the optimal filtering algorithm to obtain an output map. Window optimization angle filtering is used as a non-linear direction filtering method, the direction with the highest data consistency is searched in a sliding window through iteration, interpolation averaging is carried out in the direction, the trend feature of an anomalous body can be highlighted during effective denoising, boundary information is reserved, the method is suitable for magnetic anomaly processing with obvious directivity, and the method is suitable for large-scale popularization and application. The window optimization angle filtering can effectively enhance the trend characteristics of the anomalous body, avoids the problems of anomaly superposition and boundary smoothness, and provides a clearer and more accurate data basis for subsequent geological interpretation.

The application relates to the field of intelligent medical treatment, in particular to a thyroid ultrasound image data enhancement method, device and program product. The method comprises the following steps: acquiring a thyroid ultrasound image; segmenting the thyroid ultrasound image to obtain a fascia mask and an adenoid parenchyma mask; calculating the distance from each pixel to the nearest edge pixel of the fascia mask to generate a distance map, and calculating the weight of each pixel through the distance map to generate an anatomical weight atlas; performing smoothing processing on the thyroid ultrasound image through the anatomical weight atlas to obtain an illumination layer; removing the pixel component of the illumination layer on the pixel of the thyroid ultrasound image to obtain a texture layer; performing directional filtering on the texture layer to obtain a direction selection coefficient; performing gate control on the tissue region of the adenoid parenchyma mask to obtain a tissue gate map; enhancing the texture layer through the direction selection coefficient and the tissue gate map to obtain an enhanced texture layer; and combining the illumination layer and the enhanced texture layer to obtain an enhanced image. The application has good clinical value.

The invention relates to a backup node scheduling method and device, computer equipment, a computer readable storage medium and a computer program product. The method comprises the following steps: acquiring performance index information of each candidate backup node in a database cluster; inputting the performance index information into a directional filter to obtain a positioning filtering result; wherein the directional filter is used for directly determining a target backup node; under the condition that the target backup node is not determined in the positioning filtering result, inputting the performance index information into a condition filter to obtain a candidate node list after condition screening; and calculating a performance evaluation score of each candidate backup node in the candidate node list according to the index evaluation weight, determining a target backup node according to the performance evaluation scores, and executing a backup task by adopting the target backup node. By adopting the method, the backup node scheduling accuracy in the database cluster environment can be improved.

The invention belongs to the field of urban road network positioning and recognition, and particularly discloses a vehicle trajectory data analysis method based on a direction constraint density clustering algorithm, which comprises the following steps: preparing and processing data; constructing a spatial index; directional filtering is performed; carrying out clustering expansion; performing preliminary clustering and small cluster merging; and finally, outputting a clustering label. According to the method, the problems of high directivity and complex traffic flow direction in road traffic data are solved by introducing direction-sensitive distance measurement; by dynamically adjusting the direction of the cluster, the clustering algorithm can better adapt to complex areas such as road turning and intersections, so that the separation of bidirectional roads is effectively realized, and the defect of the existing method in processing the directivity of vehicle track data can be effectively overcome.

The invention provides an ultrasonic imaging method and device, electronic equipment and a storage medium, and the method comprises the steps: determining a basic median angle of a reconstruction point based on the relative position information between the reconstruction point on a detection target and an ultrasonic probe, the basic median angle being an included angle between the normal of the reconstruction point on the surface of the detection target and the normal of the surface of the ultrasonic probe; determining a median angle range including a basic median angle; based on echo data obtained by detecting the detection target by the ultrasonic probe, determining a brightness component of the reconstruction point under each mid-angle in the mid-angle range; and determining a brightness value of the reconstruction point in the ultrasonic image based on the brightness component of the reconstruction point at each mid-angle in the mid-angle range. According to the method, the device, the electronic equipment and the storage medium provided by the invention, directional screening and superposition of the echo signals are realized, the visual definition under various inclination angles and depths can be effectively improved, the existing ultrasonic equipment hardware and data acquisition process are prevented from being changed, and the method and the device have compatibility and real-time performance.

The invention discloses a channel fingerprint generation and recovery method based on a conditional latent diffusion model, and the method comprises the steps: converting a channel fingerprint into an initial image through a preset mapping rule, carrying out the graying verification of the initial image, carrying out the multi-scale direction filtering processing of a gray image, and obtaining a channel fingerprint image; inputting the filtered image as a condition of a conditional latent diffusion model, initializing latent variables, expanding pixel values of the filtered image into feature vectors, executing a reverse sampling process, and converting a final latent variable obtained in the reverse sampling process into data consistent with an initial channel fingerprint dimension from the final latent variable of the forward diffusion process, and the generated and recovered channel fingerprints are obtained. Therefore, the beneficial effects of optimizing the preprocessing quality to reduce the feature loss, improving the feature capture integrity to enhance the fingerprint discrimination, enhancing the generated fingerprint similarity to guarantee the feature authenticity, improving the complex environment robustness to guarantee the recovery stability, guaranteeing the dimension consistency to improve the engineering practicability and the like can be achieved.

The application belongs to the field of urban road network positioning and recognition, and specifically discloses a vehicle trajectory data analysis method based on a direction-constrained density clustering algorithm, which comprises the following steps: data preparation and processing; construction of a spatial index; directional filtering; cluster expansion; preliminary clustering and small cluster merging; and finally outputting a cluster label. The application introduces a direction-sensitive distance measurement to solve the problem of strong directionality and complex traffic flow direction in road traffic data; through dynamic adjustment of the direction of the cluster, the clustering algorithm can better adapt to complex areas such as road turns and intersections, thereby effectively separating bidirectional roads and effectively solving the defects of the existing methods in handling the directionality of vehicle trajectory data.

The invention provides an ISP chip, an image processing method, equipment and a storage medium, and the ISP chip comprises a demosaicing module which is used for determining the edge direction of each initial pixel point in an initial image; based on the edge direction of each first to-be-interpolated pixel point, performing first color channel interpolation on each first to-be-interpolated pixel point to obtain a first intermediate image; performing second color channel interpolation on each second to-be-interpolated pixel point based on the edge direction of each second to-be-interpolated pixel point to obtain a second intermediate image; performing third color channel interpolation on each third to-be-interpolated pixel point based on the edge direction of each third to-be-interpolated pixel point to obtain a third intermediate image; performing directional filtering on each first middle pixel point to obtain a target filtering value of each first middle pixel point; and superposing the target filtering value of each first intermediate pixel point and the third intermediate image to obtain a target image. According to the technical scheme, a hardware structure can be reused, and the hardware area is reduced.

The present invention belongs to the technical field of mobile robot positioning and mapping. A method and system for indoor loop detection based on angle features is provided. Keyframe point clouds are subjected to highly directional filtering to generate an average point cloud. Angle feature points are extracted from the average point cloud. Convex and concave angles are discriminated for each angle feature point. A triangle descriptor is constructed based on the angle feature points and the corresponding convex and concave angle discrimination results. A hash table structure is used to uniformly manage triangle descriptors in historical keyframes. Multiple candidate loop frames are determined based on each triangle descriptor of the current keyframe and the triangle descriptors of the historical keyframes in the hash table structure. Loop verification is performed on each candidate keyframe to determine the optimal loop frame. The two-dimensional rigid body transformation corresponding to the optimal loop frame is used as the initial pose estimate for point cloud registration between the current keyframe and the optimal loop frame, thereby achieving highly robust loop detection.

The invention provides a vehicle external sound processing method and system and a storage device. The method comprises the steps that original environment sound signals in multiple directions outside a vehicle are collected; based on a target sound pointing instruction, performing beam forming processing on the original environment sound signal, and extracting a first sound signal in a specific direction; performing noise suppression and target sound separation processing on the first sound signal to obtain a second sound signal; and performing high-fidelity audio reconstruction and amplification on the second sound signal to obtain an audio signal of the target sound, and outputting the audio signal to an in-vehicle loudspeaker. According to the method provided by the invention, directional screening and noise reduction of the environment sound outside the vehicle and high-fidelity playback in the cabin can be realized.

The application discloses an image spatial domain noise reduction method and device, the method comprises the following steps: establishing a multi-scale pyramid according to an image to be reduced, and establishing a directional filtering pyramid by using the intermediate result of the Gaussian pyramid, and fusing and reconstructing the image to be reduced according to the non-local mean filtering result of each pyramid. It can be seen that the application can effectively avoid the situation that the scene adaptability is not enough by using the multi-scale pyramid, and the problem that the texture edge of the image after noise reduction is excessively smoothed or edge color noise residues are generated is effectively avoided by adding the directional filtering pyramid, the overall definition is considered while the noise reduction effect is ensured, and the efficiency and accuracy of the image noise reduction work are improved.

The utility model discloses a wireless gun searching device based on space directional filtering, which comprises a bottom shell, a fixing plate is mounted at the upper end of the bottom shell, a middle layer mechanism is arranged on the fixing plate, a top shell is mounted on the upper side of the fixing plate, an upper layer mechanism is arranged in the top shell, and a lower layer mechanism is arranged on the upper side of the top shell. The upper layer mechanism comprises a main control board arranged in the top shell, the main control board is connected with a display screen which is arranged in the upper surface of the top shell and used for displaying gun position signals and manual operation, and a working instruction is given to the main control board by operating the display screen at the top. And controlling the steering engine to rotate, driving the space directional filter to rotate, reading a wireless signal of a corresponding gun, obtaining angle and distance information of the gun, sending data back to the main control panel, displaying a position signal of the corresponding gun on the display screen, and providing direction guidance for a user to search the gun.

The invention discloses a wireless gun searching device based on spatial directional filtering, which comprises a bottom shell, a fixed plate is mounted at the upper end of the bottom shell, a middle layer mechanism is arranged on the fixed plate, a top shell is mounted on the upper side of the fixed plate, an upper layer mechanism is arranged in the top shell, the upper layer mechanism comprises a main control board arranged in the top shell, and a wireless module is arranged in the main control board. The main control board is connected with a display screen which is arranged in the upper surface of the top shell and is used for displaying gun position signals and manual operation, and by operating the display screen at the top, a working instruction is given to the main control board, the steering engine is controlled to rotate, the space directional filter is driven to rotate, and corresponding gun wireless signals are read; the angle and distance information of the gun is obtained, the data is sent back to the main control board, the position signal of the corresponding gun is displayed on the display screen, and direction guidance for searching the gun is provided for a user.

The invention relates to the technical field of intelligent inspection of electric power facilities, in particular to an overhead transmission line airborne point cloud classification method, system and device and a medium, point cloud data of a transmission line are collected, key features are extracted by adopting a statistical method, and data integrity is optimized in combination with a subspace extension strategy and an elevation statistical histogram; constructing a geometric surface by using a mathematical method to carry out directional filtering and curved surface fitting, and filtering terrain interference and unrelated vegetation point clouds; based on a particle swarm optimization algorithm, adaptively searching linearity, curvature and elevation threshold values, and realizing coarse classification of the transmission conductor and the tower; a subspace feature extension strategy is combined with a clustering analysis algorithm, horizontal extension tower data are dynamically discriminated, a vertical extension threshold value is optimized, and fine classification is completed; the problems of incomplete point cloud extraction, complex scene misjudgment and insufficient small sample classification precision are solved, and technical support is provided for intelligent inspection of the overhead transmission line.

The invention discloses a block and frame directional filtering method in CAD drawing AI translation, and relates to the technical field of drawing translation. In the field of CAD drawing AI translation, in the prior art, due to the fact that global characters of a drawing are extracted and translated without difference, company names, logos and the like in a drawing frame and content, bearing fixed formats and identities, of project names, patent information and the like in a specific drawing block are often wrongly translated, normalization and authority of the drawing are damaged, additional manual proofreading and correction are needed, and translation efficiency is reduced. According to the method, the structured model is constructed by analyzing the CAD file, and the block index is established, so that a user is allowed to specify a specific block or frame name which needs to be reserved, and the specific block or frame name is uniformly marked as an exclusion translation state; therefore, it is ensured that core contents, such as company names, logos and signature columns, bearing fixed formats and identity information in a picture frame are completely reserved in the follow-up AI translation process, and the problem that the normalization and authority of drawings are damaged due to global character extraction and translation is solved.

The invention discloses an LED dome screen display interpolation method based on edge detection. Belongs to the technical field of LED dome screen display. The technical problem that the existing uniform point extraction technology does not solve the problem of image information loss is solved. The method comprises the following steps: S1, extracting pixel intensity variation of an input image in four directions, namely 0-degree, 45-degree, 90-degree and 135-degree directions, and generating a pixel intensity variation matrix; s2, carrying out binarization processing on the pixel intensity change matrix, generating a binary contrast guide decision diagram, and marking an edge region which needs to be subjected to directional filtering; s3, performing edge diffusion processing on the edge region needing direction filtering based on the contrast to obtain a contrast-guided edge decision diagram; and S4, performing interpolation processing on the pixel points according to the mark information of the contrast-guided edge decision diagram to obtain a final display image of the input image.

This application relates to the interdisciplinary field of geotechnical engineering and computer vision, and discloses an image recognition and analysis system for shallow foundation reinforcement conditions. The system includes an image acquisition and separation module for acquiring the foundation video stream and dividing it into a pre-reinforcement steady-state image, a dynamic sequence during reinforcement, and a post-reinforcement steady-state image; a transient displacement calculation module for solving the dynamic sequence and extracting the transient displacement field gradient mean and transient principal displacement vector angle; a frequency domain directional filtering module for mapping the principal displacement vector angle to orthogonal directions and constructing a directional two-dimensional Gabor filter bank to process the steady-state image; a spectral feature calculation module for calculating the high-frequency texture energy integral and outputting the directional compaction anisotropy index; and a state quantification evaluation module for fusing the displacement field gradient mean and anisotropy index to generate a reinforcement state index. This invention establishes a cross-domain mapping between macroscopic deformation and microscopic texture, improving the accuracy of foundation compaction assessment and realizing adaptive closed-loop control of foundation reinforcement construction.

The application discloses a channel fingerprint generation and recovery method based on a conditional latent diffusion model, comprising the following steps: converting the channel fingerprint into an initial image through a preset mapping rule, performing gray scale verification on the initial image, performing multi-scale directional filtering processing on the gray scale image, taking the filtered image as the condition input of the conditional latent diffusion model and initializing the latent variable, expanding the pixel value of the filtered image into a feature vector, performing a reverse sampling process, starting from the last step latent variable of the forward diffusion process, converting the final latent variable obtained through the reverse sampling process into data consistent with the dimension of the initial channel fingerprint, and obtaining the generated and recovered channel fingerprint. Thus, the preprocessing quality can be optimized to reduce feature loss, the feature capture integrity can be improved to strengthen the fingerprint distinguishability, the generated fingerprint similarity can be enhanced to guarantee the feature authenticity, the complex environment robustness can be improved to guarantee the recovery stability, and the dimension consistency can be guaranteed to improve the engineering practicability, and the like.

The present invention discloses an intelligent identification method for underwater dredging pipelines based on multi-temporal remote sensing images, including: obtaining multi-temporal remote sensing images of the same study area; extracting the blue band, green band and near-infrared band, or the blue band, green band and short-wave infrared band; calculating the optical remote sensing water body index and displaying it using a rainbow color map; determining the optimal threshold for water body segmentation of the optical remote sensing water body index; clarifying the water body and non-water body areas; using a Lee filter for image enhancement; using a custom convolution kernel for directional filtering enhancement; using a double-threshold sobel edge detection method to obtain an edge detection result map; superimposing the edge detection result map on the rainbow color map, patching the discontinuous parts at the middle position of the continuous line of the edge detection result map, and removing the debris non-segment edges to form a complete display map of the distribution of dredging pipelines; it can achieve efficient, large-scale scanning and precise positioning of the pipelines without disturbing the normal operation of the reservoir.

The invention provides an intelligent evaluation method and system for the scale depth of a micro puncture guide wire and a storage medium, and the method comprises the steps: obtaining continuous multi-frame images of a target region, and generating a high-signal-to-noise-ratio image through inter-frame motion estimation and time domain weighted fusion; multi-directional filtering is carried out by adopting a controllable filter bank, and a guide wire central axis and a contour boundary are extracted based on maximum linear response; extracting a one-dimensional pixel intensity sequence and a local width sequence along the axis; performing continuous wavelet transform on the intensity sequence, determining an analysis scale by combining with a local width, and extracting a scale mark initial position along a scale-space energy ridge line; constructing a nonlinear pixel-physical length correction model according to a guide wire physical diameter and width sequence, and converting an initial position into an accurate physical coordinate; and by taking any scale as a reference, calculating the physical distance between the other marks and the scale, and determining the puncture depth.

The invention discloses a circuit breaker voiceprint acquisition orientation correction method based on ultrasonic position calibration, and the method achieves the real-time space positioning of a moving contact through the deployment of an ultrasonic transmitter at the moving contact of a circuit breaker, the installation of a pre-calibration microphone array in a switch cabinet, and the triggering of an ultrasonic pulse through the switching-on and switching-off actions. On the basis of a positioning result, distance attenuation compensation and directional filtering parameters of the voiceprint signal are dynamically adjusted, acoustic characteristic deviation caused by mechanical vibration or environment temperature change is eliminated, finally, the corrected voiceprint signal is input into a fault diagnosis model, and the detection precision is improved. The method is suitable for on-line monitoring and intelligent operation and maintenance of power switch devices such as high-voltage circuit breakers and GIS equipment.

The invention provides a false liquid level identification method and system. According to the method, an original image of a liquid level area of the steam generator is collected in real time, background illumination difference is removed through graying, CLAHE contrast enhancement, bilateral filtering smoothing, Laplacian sharpening and Top-hat conversion, and an enhanced grey-scale map is obtained; then, a fractal dimension graph based on a sliding window is calculated to represent local texture complexity, background and stripe artifacts are removed through local fuzzy subtraction and directional filtering, after the edge is detected through a Sobel operator, a bubble mask is generated in combination with expansion-erosion operation and morphological closed operation, and edge noise is eliminated through edge reverse masks; and finally, calculating the space overlapping degree of the liquid level edge and the bubble mask, judging the liquid level as a false liquid level when the overlapping degree exceeds a threshold value, and otherwise, judging the liquid level as a real liquid level. The method and the system have the advantages of high identification precision, short response time, strong steam interference resistance and the like, and are suitable for liquid level monitoring and safety early warning in high-temperature and high-pressure environments of nuclear power plants, thermal power plants and the like.