10 results about "Transform coding" patented technology

This invention discloses a point cloud attribute encoding method, apparatus, decoding method, and apparatus. The point cloud attribute encoding method includes: sorting all point cloud data to be encoded to obtain sorted point cloud data, wherein the point cloud data to be encoded is point cloud data with attributes to be encoded; constructing a multi-layer structure based on all sorted point cloud data and the distances between each sorted point cloud data; obtaining the encoding method corresponding to each node in the multi-layer structure, wherein the encoding method corresponding to a node is a direct encoding mode, a predictive encoding mode, or a transform encoding mode, wherein the predictive encoding mode encodes the node based on information from its neighboring nodes, and the transform encoding mode encodes the node based on a transform matrix; and performing point cloud attribute encoding on each node based on the multi-layer structure and the corresponding encoding method. Compared with existing technologies, this invention improves the overall encoding efficiency of point cloud data.

The application discloses a kind of based on twin architecture optical communication device surface defect detection network structure and detection method, belong to image processing related technical field.The surface defect detection network structure includes feature extraction encoder and feature fusion decoder;Feature extraction encoder is composed of twin ResNet residual network, non-defect feature matching elimination module and defect feature enhancement module;Defect feature enhancement module is composed of Transform coding decoder and convolution triplet attention module;Feature fusion decoder uses multilayer perception module.Improved ResNet residual network is used to extract the multi-scale image features of the sample image to be tested and the template sample image simultaneously, eliminate non-defect features, obtain difference feature maps, and output the segmentation results of the final defect area after defect feature enhancement and multilayer perception processing.The application realizes accurate and rapid segmentation of the surface defect area of the optical communication device, significantly improves the efficiency and accuracy of the surface defect segmentation and defect detection of the optical communication device.

A method of decoding can be performed by at least one processor and can include receiving an entropy encoded bitstream including compressed video data, generating one or more dequantized blocks, determining whether at least one of a height and a width of the one or more dequantized blocks is greater than or equal to a predetermined threshold, and in response to at least one of the height or the width of the one or more dequantized blocks being greater than or equal to the predetermined threshold, using a tuned look-up-table transform (LGT) kernel to transform encode the dequantized blocks to perform a direct matrix multiplication for each of a horizontal dimension and a vertical dimension of the one or more dequantized blocks.

A method of decoding may be performed by at least one processor, and may comprise: receiving an entropy coded bitstream comprising compressed video data; generating one or more dequantized blocks, determining whether at least one of a height and a width of the one or more dequantized blocks is greater than or equal to a predefined threshold, and responsive to the at least one of the height or the width of the one or more dequantized blocks being greater than or equal to the predefined threshold, transform coding a dequantized block using a tuned line graph transform (LGT) core to perform direct matrix multiplications for each of the horizontal and vertical dimensions of the one or more dequantized blocks.

Disclosed herein are systems, methods, and instrumentalities associated with transform coding. Candidate transforms for a video block may be ordered by a video encoder based on respective hypothetical costs associated with processing the video block based on the candidate transforms. A transform index value indicating the position of a suitable transform on the ordered transform list may then be signaled to a video decoder, which may perform similar cost-based sorting operations to derive the ordered transform list and select a transform for the video block from the ordered transform list based on the signaled index.

The application discloses a text similarity calculation method based on scene information enhancement, comprising the following steps: obtaining two texts, generating scene posterior and target label and calculating consistency coefficient; using pre-training Transform coding and applying scene gating to obtain two scene weighted embeddings; performing scene modulation and counterfactual construction, combining intervention invariance and adversarial debiasing training to obtain debiased embedding; calculating weighted Jaccard lexical difference according to the target scene; fusing lexical difference and semantic distance to construct cost, solving transmission plan by entropy regularization to obtain scene-coupled optimal transmission distance; calculating semantic similarity and fusing conditional value at risk to obtain robust semantic similarity; mapping the transmission distance to alignment similarity, and weighting the robust semantic similarity to output comprehensive similarity. The application improves the accuracy and robustness of cross-scene similarity determination.

The present application relates to a kind of encoding method and system of orthogonal polygon distribution mode, the method includes: first establish polygon database and mode database, then find every polygon in the graph to be encoded in polygon database, find success then return the number and transformation type, otherwise according to the shape of polygon encoding judge its symmetry type, and find all its transformation encoding, store in the polygon database and assign a polygon number, after returning the polygon type number and transformation type, then according to polygon number and transformation type and its relative to the position of mode center point encode mode, finally create various transformation encoding of mode.By polygon-based, very friendly to realize encoding, without image or grid conversion, graph or mode symmetry processing and conversion is convenient, simple, also can handle polygon overlap and the case of hole-containing polygon, and graph and encoding conversion is convenient, subsequent operation data volume is small.

This invention relates to the fields of remote sensing image processing and agricultural and forestry monitoring technology, and provides a method and apparatus for compressing spectral data from unmanned aerial vehicles (UAVs). The method includes: acquiring spectral image data of a target area and determining a set of multi-indicator functional traits to be inverted in the target area; calculating the joint importance score of each spectral band to all functional traits based on the set of multi-indicator functional traits using a spectral sensitivity analysis model; dividing each spectral band into a core region, a transition region, and a redundant region according to the joint importance score; employing deep learning-based feature-preserving encoding for the core region, transform encoding for the transition region, and feature extraction encoding for the redundant region; and decoding and reconstructing the encoded streams of each region to obtain the target spectral data used for functional trait inversion. This achieves maximum preservation of key spectral information required for multi-indicator collaborative inversion under high compression ratio conditions, ensuring the accuracy of functional trait inversion in the reconstructed data.

The present application relates to the technical field of pipeline safety monitoring, and particularly relates to a gas pipeline monitoring method based on vibration response space-time feature fusion modeling, comprising: collecting the multi-measurement-point vibration response of a low-pressure gas pipeline, and then performing spatial combination and standardization pretreatment on the multi-measurement-point vibration response; performing time-dimension Transform coding; performing space-dimension Transform coding; performing space-time feature fusion based on gate fusion; and performing two-stage vibration classification driven leakage identification.The present application faces the multi-measurement-point vibration response signals of adjacent points, simultaneously extracts the evolution features in the time dimension and the propagation correlation features in the space dimension, and realizes feature fusion through an adaptive mechanism; meanwhile, in combination with a layered discrimination idea, the screening of normal state and abnormal state is realized first, and then the abnormal event type is further distinguished, so that the reliability and engineering applicability of the weak leakage identification of the low-pressure gas pipeline in a complex residential environment noise background are improved.