7 results about "Gabor filter" patented technology

The application relates to a bridge structure modal parameter identification method, which is based on a bridge structure micro-vibration video collected by a camera, fuses a variational mode decomposition (VMD), a phase-based video motion magnification (PBVM), a Gabor filter-based phase extraction method (GBP), an improved clustering algorithm (ICA) and an integrated covariance-driven stochastic subspace identification (SSI-Cov) and frequency domain decomposition (FDD) algorithm, and constructs a complete non-contact structure modal parameter identification system. In the absence of prior knowledge of the target frequency bandwidth, environmental noise interference and without obvious characteristic targets, the effective decoupling of multiple vibration modes of the bridge structure, the frequency domain feature extraction of small-amplitude vibration and the intelligent and automatic modal parameter extraction are realized, and the application belongs to the technical field of bridge structure health monitoring.

This application provides an anti-interference texture enhancement localization method, system, and medium suitable for low-texture visual scenes. The method includes: acquiring an original image; preprocessing the original image; extracting the texture response of the preprocessed image in different directions and frequency dimensions based on multi-scale Gabor filtering to obtain multiple texture response maps; dividing the preprocessed image into multiple image partitions based on a partitioned LED array and calculating the local gradient contrast of each image partition; independently adjusting the light source brightness of different image partitions to obtain an illumination-optimized image; performing feature fusion processing on the multiple texture response maps and the illumination-optimized image to extract feature points for visual localization; improving the consistency of images acquired at different time periods by normalizing the frequency response of the original image, histogram matching, and adaptive brightness restoration; and improving the local contrast by using a partitioned LED array with spatial illumination modulation to improve visual localization accuracy.

The application discloses a kind of based on adaptive dynamic fusion YIQ color space acquisition vibration displacement realizes bearing fault feature extraction method, comprising: the YIQ color space conversion of each frame image in two-dimensional image sequence is carried out, obtains luminance component Y, in-phase chrominance component I and quadrature chrominance component Q and carries out normalization processing;In region of interest, the statistical characteristics of three normalized components are calculated respectively;According to the statistical characteristics of Y, I, Q three normalized components in region of interest, calculate dynamic fusion weight, and according to dynamic fusion weight, three normalized components are weighted fusion, and obtain the image of fused region of interest;The fused region of interest image is carried out smooth processing, obtain the smoothed fusion region of interest image and using multi-wavelength, multi-direction Gabor filter is convolved, obtains corresponding local amplitude and local phase, to obtain the continuous vibration displacement signal of target bearing in this way and then carries out fault feature extraction.The application breaks through the limitation that based on vibration displacement is only applicable to single fault type, and envelope spectrum frequency multiplication feature extraction is not complete when multiple faults occur, realizes the stable extraction of 1 frequency multiplication, 2 frequency multiplication and other key frequency multiplication components under the condition of inner ring, outer ring fault and composite fault and other multiple fault types.

This invention discloses a remote sensing image semantic segmentation method enhanced by learnable constrained convolution. The method includes: preparing a remote sensing image dataset; constructing an encoder-decoder network; in the first stage, replacing the deep standard convolutional layers of the encoder with learnable constrained convolutional modules, training for 50 epochs at a learning rate of 2×10⁻⁵; in the second stage, replacing them back with standard convolutional layers and continuing training for 100 epochs. The learnable constrained convolutional modules include Laplacian normalization units and six structural prior units (edge ​​detection, separable structure, horizontal symmetry, Gabor filtering, Gaussian smoothing, and channel antagonism); and finally, calling the final model to complete the semantic segmentation of the remote sensing image. This invention also proposes an LCMK decoding module that integrates a 2D state-space model with a Kolmogorov-Arnold network.

The rotor core for an embedded magnet synchronous motor, which includes a magnet arrangement region and a flux barrier region, can be formed to be smooth overall while having a fine slit shape. [Solution] A rotor core design method for designing the shape of a rotor core comprising a magnet arrangement region and a flux barrier region involves arranging multiple Gaussian bases (52) in the design target region (50), defining a shape function by superimposing Gabor filters at wavelengths individually set for each of the multiple Gaussian bases (52), and using the shape function, determining a core cross-sectional shape that meets pre-set objective conditions using an optimization algorithm.

This invention discloses a real-time aerial small target detection method based on phase awareness and frequency domain enhancement. First, a spatial-frequency dual-stream decoupling module simultaneously captures the target's local spatial semantics and high-frequency gradient structure during the feature extraction stage, and a hard-coded structural prior is constructed using the Scharr operator. Then, an absolute texture feature is extracted using a low-level frequency enhancer combined with a Log-Gabor filter. Based on this, a semantically gated phase interaction mechanism is used to strip away the easily disturbed amplitude spectrum in the frequency domain and reconstruct the structural prior using pure phase features. This is combined with a high-level semantic mask to achieve strong noise resistance across scales. Finally, at the detection head output, a frequency domain edge-gated hybrid expert module based on the Laplacian operator is used to achieve pixel-level dynamic feature refinement. This invention effectively revives sub-pixel-level target details and suppresses background false alarms, significantly improving the recall rate of aerial small target detection while maintaining extremely high real-time inference efficiency.

The application discloses a kind of engineering machinery fingerprint biological recognition man-machine binding attendance methods and systems, comprising: by the block calculation and correction of pressing stability degree to fingerprint image, dynamically generate block adaptive filter weighting coefficient, then combine adaptive two-dimensional Gabor filter and enhance fingerprint image block by block, effectively solve the differentiation blur and ghost interference problem generated by uneven pressing and vibration sliding of fingerprint in strong vibration environment of engineering machinery, overcome the defects that traditional global fixed filter intensity is easy to cause clear area excessive filtering, fuzzy area enhancement is insufficient and callus, oil stain interference leads to low recognition accuracy, man-machine binding attendance is unreliable, significantly improve fingerprint recognition success rate and authorized reliability, meet the safety supervision demand of engineering machinery construction site.