61 results about "Lossless coding" patented technology

The invention relates to a three-dimensional reconstruction-oriented Gaussian point cloud lossless coding and decoding method, a product and a storage medium. The method comprises the following steps: extracting a first Gaussian point cloud corresponding to anhor data in a numpy array format at a coding side; converting the first Gaussian point cloud in the numpy array format into a first Gaussian point cloud in a ply format; meanwhile, through a Morton code-based spatial sorting method, performing Morton code sorting on geometric data and attribute data of the first Gaussian point cloud in the numpy array format, and generating a Morton sequence index table; and carrying out AVS coding on the first Gaussian point cloud in the ply format to obtain compressed data in a binary code stream form. On the decoding side, AVS decoding is carried out when the compressed data in the binary code stream form and the Morton sequence index table are received, and a second Gaussian point cloud in the ply format is obtained through reduction; performing format conversion on the second Gaussian point cloud in the ply format to obtain a second Gaussian point cloud in a numpy array format; and rearranging the attribute data and the geometric data based on the Morton sequence index table to realize one-to-one correspondence of the geometric data and the attribute data between the first Gaussian point cloud and the second Gaussian point cloud before and after coding and decoding. Compared with an existing method, the data consistency before and after coding is improved.

The invention relates to an array microphone noise reduction recording method based on cascade noise reduction and blind source separation, which belongs to the technical field of voice signal processing and recording, and comprises the following steps: configuring a multi-channel array microphone, ensuring that the amplitude and phase of a channel signal are consistent, and collecting an original multi-channel voice signal; a weighted kernel function blind source separation algorithm is adopted, signal-to-noise ratio distribution characteristics of signals are extracted, kernel function weights are given, and target voice and interference signal components are obtained through decoupling of an independent component analysis model; executing target-oriented adaptive cascade noise reduction, locking the voice of a keynote speaker through directional pickup, reducing noise, filtering out reverberation, and enhancing the voice of a far-field target by combining a voice mask neural network with a far-field pickup algorithm in sequence; and processing the target voice through voice feature perception lossless coding and storing the target voice. According to the invention, stable acquisition of multi-channel signals, accurate separation of mixed signals and layered suppression of noise reverberation are realized, the signal-to-noise ratio and definition of far-field voice are significantly improved, and the method is suitable for single-person speaking or multi-person dialogue scenes.

The invention discloses a GPU shader rendering computer implementation method of a finite element result. The method comprises the steps that a static three-dimensional grid is adopted as a rendering geometry; the 32-bit floating point type scalar data corresponding to the vertexes are subjected to lossless coding at a CPU end to obtain four-channel 8-bit RGBA vertex color attributes; during data updating, lightweight vertex color data are only transmitted to the GPU; hardware interpolation is carried out on the encoded color attributes by using a GPU rasterizer; and finally, in the fragment shader, decoding the interpolation result of each pixel to reconstruct a scalar value, and mapping the scalar value into a final color. According to the method, the bottleneck of topological calculation of a CPU end and transmission of massive geometric data to the GPU is avoided, the calculation load is transferred to the GPU in a large scale for parallel processing, the rendering efficiency is remarkably improved, and high-frame-rate dynamic visualization of a finite element result is realized.

A method and an apparatus for switching between a lossy codec and a lossless codec are provided. The method includes: obtaining a waveform of a previous frame Ti−1, and updating the waveform of the previous frame Ti−1 to an input frame buffer in the lossless codec, where the waveform of the previous frame Ti−1 is encoded by a lossy encoder; performing integer window time domain aliasing cancellation INT winTDAC on a waveform in a lossless encoder buffer to obtain a first transform result, and updating the first transform result to an overlapped buffer in the lossless codec; obtaining a waveform of a current frame Ti, and updating the waveform of the current frame Ti to the input frame buffer; and performing integer modified discrete cosine transform INTMDCT on the waveform in the lossless encoder buffer to obtain a second transform result.

The embodiment of the invention relates to an image encoding apparatus and an image decoding apparatus. An encoded semantically segmented image of an input image is generated based on a semantic segmentation deep neural network (120) and a lossless encoder (130). An encoded coarse image of the input image is generated based on a first lossy encoder (140) and the semantic-based distortion map. An estimated image of the input image is generated based on a semantic segmented image in a fine deep neural network (150), a coarse image, and the semantic-based distortion map. Based on the estimated image, the input image and a second lossy encoder (160) generate an encoded residual image. Furthermore, embodiments of the invention also relate to a first communication device and a second communication device, and also to corresponding methods and computer programs.

Disclosed by the present invention is a lossless coding and decoding method for reducing a video transmission bandwidth, thus improving real-time performance of a system. In addition, A floating-point number from probability calculation is prevented from affecting the coding. Two-stage compression improves the compression ratio with simple and practical operations. The present invention greatly reduces the bandwidth resources occupied in the video transmission process, thus having extremely high application value in the occasions with requirements for high-quality video image.

Methods and apparatuses are provided for video coding with lossless coding modes. The method includes: determining whether a current block is in a transform skip mode; in response to determining that the current block is in the transform skip mode, determining whether a first residual coding scheme or a second residual coding scheme is selected for the current block in the transform skip mode; and signaling a first indication which indicates whether a current block is in the transform skip mode and signaling a second indication which indicates whether the first residual coding scheme or the second residual coding scheme is applied for blocks in the transform skip mode.

The decoder includes circuitry configured to receive a bitstream in which a current frame is identified, wherein the current frame includes a first region and a third region, detect in the bitstream an indication that the first region is encoded according to a lossless encoding protocol, and decode the current frame, wherein decoding the current frame further includes decoding the first region using a lossless decoding protocol corresponding to the lossless encoding protocol.

The application provides a mass medical image cold data joint compression method, a feature extraction network is used for fine-grained feature extraction on large-scale medical images, a feature database of medical images is established, and the extracted image features are stored according to classification; the feature extraction network is used for the medical image to be compressed, features are extracted and similarity matching is performed; then, based on image brightness, slices of the reference image and the image to be compressed are sorted to optimize the compression order; finally, according to the reference image, lossy coding of the current image is assisted, and according to the original image, lossless coding of the current image is assisted.

The invention discloses a high-precision wave recording acquisition and time mark synchronization system for power grid fault positioning, and relates to the technical field of power grid wave recording synchronization, and the system comprises the steps: 1, synchronously outputting three-gear sampling under a time axis of the same coordinated world time, and generating a time uncertainty upper bound and a group delay signature for a channel and a fragment; 2, performing quality locking near-lossless on the high-energy fragment, after coding, checking the high-energy fragment to reach a sample position and a peak amplitude through homologous detection, and returning to be lossless if the high-energy fragment is unqualified; 3, writing a fragment index, quality metadata, a fragment verification structure and a whole file signature into the compatible single file expansion container; step 4, carrying out progressive retrieval, breakpoint resume and cross-station de-duplication according to indexes, and implementing cold and hot hierarchical storage; and 5, based on group delay signature compensation and quality weighted loading, solving a reference moment and generating an acceptance check list. According to the scheme, cross-device alignment, reproducible key features, transmission and storage peak clipping, playback and evidence obtaining caliber unification and tool compatibility and auditing are realized.

The invention discloses an image coding and decoding display method and system, relates to the technical field of electronic information, and can determine an image needing to additionally transmit a quality enhanced data packet by additionally considering the operation logic of a user on a cloud desktop terminal side and the refreshing frequency of the image in a cloud desktop image coding process, thereby ensuring the user experience. According to the specific scheme, a server side encodes a to-be-displayed image in an H264 lossy encoding mode and issues the to-be-displayed image to a cloud desktop terminal to be decoded and displayed. Meanwhile, a time threshold value T is set, and under the condition that the cloud desktop image is not refreshed after the time T, the server side constructs an image quality enhancement data packet of lossless coding and sends the image quality enhancement data packet to the cloud desktop terminal to be decoded and displayed. The server side sets a time threshold T according to the user input quantity uploaded by the cloud desktop terminal, and the larger the user input quantity is, the smaller the T is.

The application discloses a point of interest check-in sequence generation method based on a diffusion model, and belongs to the technical field of spatiotemporal data mining and deep learning. Data preprocessing is performed to clean and smooth a real world data set, so as to ensure data quality and consistency. Spatiotemporal lossless coding is performed to convert check-in sequences of different lengths into check-in vectors of equal lengths, and the original sequence information is retained by coding through a spatial frequency vector and a time bucket vector. A diffusion model is established to construct a spatial diffusion module and a time diffusion module, and spatiotemporal features are captured by using a forward diffusion process and a backward reconstruction process. A conditional U-shaped network is introduced into the diffusion module, and a denoising network is proposed to capture complex spatiotemporal correlations by modeling through a self-attention mechanism. A contrast learning strategy is used to further capture the spatiotemporal correlations of the check-in sequence by using ternary contrast learning, and the connection between the time and space diffusion modules is strengthened.

The invention provides an image coding transmission method and device based on a cache mechanism, equipment and a medium, relates to the technical field of computers, and solves the problem that the transmission bandwidth is wasted due to the fact that the same image content is coded for multiple times in the prior art. According to the scheme, the macro blocks subjected to lossless coding can be stored at the coding side by utilizing the cache mechanism, so that the macro blocks with the same image content do not need to be coded again in subsequent processing, the coding times of the same image content are effectively reduced, the coding efficiency of the image is further improved, and the coding efficiency of the image is improved. The code rate of cloud desktop hybrid coding is reduced, and the effect of saving bandwidth is achieved.

An example device includes a memory configured to store video data; and one or more processors implemented in circuitry and coupled to the memory. The one or more processors are configured to determine whether a transform skip mode is used for a current block of the video data. The one or more processors are configured to disable level mapping for residual coding based on the transform skip mode being used for the current block. The one or more processors are configured to code the current block without applying the level mapping.

A method for coding a video sequence is provided that includes encoding a portion of a picture in the video sequence in lossless coding mode, and signaling a lossless coding indicator in a compressed bit stream, wherein the lossless coding indicator corresponds to the portion of a picture and indicates whether or not the portion of the picture is losslessly coded. A method for decoding a compressed video bit stream is provided that includes determining that lossless coding mode is enabled, decoding a lossless coding indicator from the compressed video bit stream, wherein the lossless coding indicator corresponds to a portion of a picture in the compressed video bit stream and indicates whether or not the portion of the picture is losslessly coded, and decoding the portion of the picture in lossless coding mode when the lossless coding indicator indicates the portion of the picture is losslessly coded.

The invention provides a memory-efficient real number field in-situ fast Fourier transform method and system, and the method comprises the steps: obtaining a real number input tensor to be processed, the dimension of the real number input tensor being r; performing forward fast Fourier transform on the real number input tensor to obtain frequency domain data; a real number buffer area with the size consistent with the dimension size of the real number input tensor is configured, the index range of the real number buffer area is 0-r-1, and storage is carried out based on the coefficient category of the frequency domain data; on the basis of the storage operation, complete frequency domain information of the real number input tensor is losslessly coded in a real number buffer area with the size of r, and the frequency domain information storage space and the real number input data space are consistent in size. The technical problems that in the prior art, the rFFT technology is inconsistent in memory space in deep learning application, bit calculation cannot be truly achieved, and modern data type support is lacked are solved.

A computing device performs a method of decoding video data by receiving a bitstream corresponding to video data encoding one or more pictures, wherein the one or more pictures are recursively partitioned into a plurality of groups at different partition levels; determining a first syntax element that corresponds to a first group of the plurality of groups; in accordance with a determination that the first syntax element has a first value: decoding the first group in a first mode and in accordance with a determination that the first syntax element has a second value: determining a respective second syntax element for each of one or more second groups within the first groups, wherein the one or more second groups are at a second partition level below the first partition level and decoding each of the one or more second groups in accordance with a value of the respective second syntax element.

When encoding a data array in a lossless manner, the data array is encoded using a lossless encoding scheme, and using an auxiliary encoding scheme that can encode at least some data arrays in a lossless manner such that the original array of data elements can be reconstructed from the encoded data. The result of encoding the data array using the auxiliary encoding scheme is then used to determine whether to use a result of the encoding of the data array using the auxiliary encoding scheme for the output encoded data array or whether to use a result of the encoding of the data array using the lossless encoding scheme for the output encoded data array.

The application provides a lossless coding method, comprising the following steps: S1, providing quantized 4X4 pixel blocks, each pixel block comprising a plurality of pixels, each pixel corresponding to a quantized coefficient; S2, predicting a Golomb-Rice parameter m value of a current pixel by a forward average value of quantized coefficients of all pixels; S3, dividing the quantized coefficient corresponding to the current pixel into a coding prefix and a coding suffix according to the Golomb-Rice parameter m, performing unary coding on the coding prefix according to a unary coding table, and performing binary coding on the coding suffix; and S4, repeating steps S2-S3 until the coding of all pixel blocks is completed. In the lossless coding method provided in the embodiment of the application, the Golomb-Rice parameter m is increasing or decreasing along with the arrangement of the pixel blocks or the pixels, and the quantized Golomb-Rice parameter m is not fixed, so that the convergence speed of coding is accelerated.

The invention discloses an image compression method and device based on hierarchical prediction and time-frequency coupling fast decision, and the method comprises the steps: firstly segmenting an input image into image blocks with the same size, respectively calculating all gradient components of each image block, calculating the gradient entropy H (G) of the image block through a weighted fusion strategy, and carrying out the calculation of the gradient entropy H (G) of the image block; dividing the image blocks into simple, medium or complex areas according to a preset threshold value, and allocating prediction modes with different calculation complexity to different areas; step-by-step calculation is carried out according to a simple, medium and complex sequence by adopting a calculation multiplexing strategy, and a later-stage prediction directly multiplexes an obtained intermediate result of a previous stage; and finally, for all candidate modes, adopting a time-frequency coupled joint distortion evaluation model, and quickly selecting an optimal prediction mode by weighting and fusing time domain and frequency domain components. According to the method, on the premise that the coding quality is lossless, the calculation complexity and the processing delay of intra-frame prediction are remarkably reduced, the overall throughput rate is improved, and the method has good universality and transportability.

This application proposes a video stream cloud-edge collaborative analysis method and device. The method includes: an FPGA edge computing device acquiring a video stream captured by a camera device; the FPGA edge computing device performing real-time key region detection on the video stream using an internally deployed lightweight neural network model to obtain key region detection results and extracting low-resolution feature vectors of the key regions; the FPGA edge computing device, through its built-in video encoding and decoding unit, performing lossless encoding and high bit rate compression on the key region video stream based on the key region detection results, and performing Gaussian blur preprocessing and high quantization parameter compression on the non-key region video stream to obtain compressed video data; and uploading the compressed video data, key region location metadata, and low-resolution feature vectors to a cloud GPU server. This can improve real-time performance and network efficiency.

Methods and apparatus are provided for video coding with lossless coding modes. The method includes: dividing a video image into a plurality of CUs comprising a lossless CU; determining a remaining coding block size of the lossless CU; and upon determining that the remaining coding block size of the lossless CU is greater than a predefined maximum value, splitting the remaining coding block into two or more remaining blocks for further coding.

A method and apparatus for video coding. Methods and apparatuses for video coding with lossless coding mode are provided. The method includes partitioning a video picture into a plurality of CUs including a lossless CU, determining a residual coding block size of the lossless CU, and in response to determining that the residual coding block size of the lossless CU is greater than a predefined maximum value, partitioning the residual coding block into two or more residual blocks for residual coding.

A method and apparatus for video encoding / decoding in a lossless encoding / decoding mode are provided, the method including: partitioning a video image into a plurality of CUs including a lossless CU; determining a residual encoding / decoding block size of the lossless CU; and, in response to determining that the residual encoding / decoding block size of the lossless CU is greater than a predetermined maximum value, splitting the residual encoding / decoding block into two or more residual blocks for residual encoding / decoding.

An electronic device performs a method of decoding video data, the method comprising: if adaptive color transform (ACT) is enabled for a residual of a coding unit, receiving, from information in a bitstream, an adjustment value applied to a quantization parameter for inverse quantization, and a coefficient for the residual of the coding unit; reconstructing the residual of the coding unit by applying inverse quantization and inverse transform to the reception coefficient for the residual of the coding unit; and, if the residuals of the coding units have been encoded using ACT, modifying the reconstructed residuals of the coding units by applying inverse ACT to convert the reconstructed residuals of the coding units from the YCgCo color space to the original color space based on the same inverse ACT matrix as one applied to one coding unit encoded in the lossy or lossless encoding mode.

The application provides a lossless coding and decoding method, device and medium for distributed audio and video processing, comprising: generating a preliminary node allocation scheme according to a frequency band grouping boundary, calculating the frequency band processing load of each node by using a load balancing algorithm, re-distributing the frequency band carried by the node whose load exceeds a preset load threshold to a low-load node to obtain a node allocation scheme; performing frequency band splitting on multi-channel audio according to the node allocation scheme, driving each node to independently encode the allocated frequency band to obtain a node frequency band signal; performing cross-node phase synchronization detection on each node frequency band signal, extracting a node pair with different phases and the corresponding frequency band, re-distributing the frequency band to a phase-coherent node to obtain an optimized node allocation scheme; performing distributed coding on each node frequency band signal by using the optimized node allocation scheme and recombining, extracting the spatial sound field positioning of the recombined signal, comparing with the original signal positioning, and identifying the final lossless audio signal meeting the preset positioning range.

An aspect refers to a coder for coding a waveform signal, comprising a zero coder and lossy coder. The zero coder is configured to code the waveform signal using zeroed LSB elimination technique; the lossy coder configured to perform lossy transform coding of a derivative of the waveform signal. Another aspect refers to a hybrid coder configured for lossy coding a signal and lossless coding of the signal, comprising: signal generator or extractor being configured to use codec parameters comprising information concerning the lossy transform coding and information concerning the lossless coding; wherein for the information concerning the lossy transform coding and information concerning the lossless coding the same codebook is used.

The invention relates to the technical field of video compression coding, and discloses an intelligent lossless video compression optimization method and device, equipment and a medium. According to the method, initial content description of an original video stream is extracted through a visual description network, content marking points are marked, and a dynamic content topological graph is generated according to the space-time relation of the content marking points. And performing semantic aggregation on the topological graph to form a semantic aggregation unit associated with a compressed semantic tag. A candidate compression strategy is matched for each unit, an initial strategy mapping table is generated, and a final strategy mapping table is screened out according to global constraint iteration. And configuring encoder parameters according to the table to carry out lossless encoding, and collecting feedback information to adjust strategies and parameters on line. According to the method, the structured association of the video content can be deeply understood, and the dynamic closed-loop optimization of the coding strategy is realized, so that the overall efficiency and the adaptive capability of lossless video compression are improved.

The invention provides a mixed image lossless coding and decoding method and system based on frame-level rearrangement, and relates to the technical field of image processing, and the coding method comprises the steps: obtaining a mixed image; performing block preprocessing on the mixed image to obtain a plurality of sub-blocks; based on a manual design feature and a learning feature, classifying the plurality of sub-blocks, and determining a classification result; according to the classification result, carrying out frame-level rearrangement on the mixed image to obtain a rearranged image; and carrying out lossless coding on the rearranged image through a lossless coding algorithm to generate a code stream containing frame-level rearranged information. Lossless decoding of the mixed image is realized, the coding efficiency of the ultra-high-definition mixed image is effectively improved, the complexity increase is less, the method is compatible with an open source encoder, and the method is easy for practical application.