86results about How to "Reduce error propagation" patented technology

Context-based adaptive arithmetic coding and decoding methods and apparatuses with improved coding efficiency and video coding and decoding methods and apparatuses using the same are provided. The method for performing context-based adaptive arithmetic coding on a given slice in an enhancement layer frame of a video signal having a multi-layered structure includes steps of resetting a context model for the given slice to a context model for a base layer slice at the same temporal position as the given slice, arithmetically coding a data symbol of the given slice using the reset context model, and updating the context model based on the value of the arithmetically coded data symbol.

Predictive encoding methods, predictive encoders, and digital systems are provided that encode input frames by computing quantized predictive frame parameters for an input frame, recomputing the quantized predictive frame parameters wherein a previous frame is assumed to be erased and frame erasure concealment is used, and encoding the input frame based on the results of the computing and the recomputing. In embodiments of these methods, encoders, and digital systems, two phase codebook search techniques used in the encoding process are provided that compute the predictive parameters in the first phase, and the predictive parameters assuming the prior frame is erased in the second phase. In the second phase, a frame erasure concealment technique is used in the computation of the predictive parameters.

The invention discloses a mechanical arm hand-eye calibration method based on the spatial distance and epipolar line constraint. The mechanical arm hand-eye calibration method comprises the steps thatan initial value of a hand-eye conversion matrix is solved, a conversion relationship of a camera fixed to a mechanical arm when the mechanical arm is at the two different positions is obtained, epipolar line errors of feature points in images are calculated through internal parameters of the camera, the feature points are subjected to three-dimensional reconstruction, the spatial distance between the reconstructed feature points is calculated, spatial distance errors are calculated according to the actual spatial distance, the cumulative sum of all the epipolar line errors and the spatial distance errors of all the feature points serve as an optimized objective function, iterative optimization is conducted through a Levenberg-Marquardt algorithm, and the hand-eye conversion matrix with higher precision is obtained. According to the mechanical arm hand-eye calibration method, the optimized objective function which better meets the actual demands can be constructed, thus the spatial size calculated by the hand-eye conversion matrix is more consistent to the actual size, the convergence rate of iterative optimization is increased, it is ensured that the optimized parameters meet theepipolar geometry constraint, and the precision is higher.

A method of significantly reducing error propagation due to voice packet loss, while still greatly profiting from long-term pitch prediction, is achieved by adaptively limiting the maximum value of the pitch gain for the first pitch cycle within one frame. A speech coding system for encoding a speech signal, wherein said a plurality of speech frames are classified into said a plurality of classesdepending on if the first pitch cycle is included in one subframe or several subframes. The pitch gain is set to a value significantly smaller than 1 for the subframes covering first pitch cycle; wherein the pitch gain reduction is compensated by increasing the coded excitation codebook size or adding one more stage of excitation for the subframes covering the first pitch cycle.

The invention discloses a driver fatigue state level classification method, belonging to the technical field of automobile safety driving assistance. The invention accurately classifies the level of fatigue state on the basis of considering the dynamic generation characteristics of driving fatigue, and has an important practical support function for related technical research and vehicle-mounted system development. This method uses reliable EEG data as the data source of fatigue level classification to reduce the error transmission caused by indirect detection and subjective evaluation methods. The level classification model based on LSTM network solves the problem of dynamic fatigue generation from time series. Attention mechanism is introduced into the model to analyze the differences ofthe characteristics under different fatigue levels, so as to improve the rationality of the model's interpretation of fatigue evolution law. The modeling method synthesizing the characteristics of different people overcomes the problem of model adaptability expansion caused by the difference of driver style types.

The invention discloses a dark network user behavior detection method and system based on network traffic, and relates to the field of computer network security. Analysis and research are carried outon user behaviors in four mainstream dark networks of Tor, I2P, ZeroNet and Freenet. User behaviors are divided into eight classes, and a hierarchical classification model is trained by establishing adark network data collection environment, collecting real dark network flow data, extracting flow characteristics in the data. According to the method, model training is carried out by adopting a hierarchical classification method, and the method is good in performance in dark network traffic identification scenes of four dark network types and 25 user behaviors and has relatively high accuracy.Meanwhile, the base classifiers in the hierarchical classifiers are independent of each other and do not interfere with each other, so that the method and the system have good expandability and can adapt to dark network environments with different ages. The system can be deployed at a flow outlet of an actual production environment and linked with an IDS and a firewall, and a solution is providedfor control and traceability of a dark network.

The present invention discloses a V-BLAST detection method for a MIMO system, including: a first detecting step for acquiring in sequence the old detection components of a transmitting symbol vector with the optimum ordering-successive interference cancellation detection method; a second detecting step for taking the detection component with the highest space diversity degree among the old detection components as a first new lo detection component, and acquiring in sequence the other new detection components of the transmitting symbol vector in an inverse order of the optimum detection order with the optimum ordering-successive interference cancellation detection method; an outputting step for outputting the first new detection component and the other new is detection components as the final detection result. The method according to the invention employs the outer close loop iteration technology to further reduce inter-layer error propagation of V-BLAST. This method has less complexity.

The embodiment of the invention provides a video data input method comprising the following steps: acquiring to-be-processed video frame data; segmenting the to-be-processed video frame data into a plurality of independent to-be-compressed blocks, and numbering the to-be-compressed blocks; scanning each to-be-compressed block according to the serial number, grouping pixel points in each to-be-compressed block while scanning each to-be-compressed block, thereby obtaining groups with different priorities; setting compression ratios for the to-be-compressed blocks; for the pixel points of different groups in one to-be-compressed block, orderly performing compression with the corresponding compression ratio according to a descending order of the priorities of the groups to generate a compressed code stream; and inputting the compressed code stream, the serial number of the to-be-compressed block corresponding to the compressed code stream and the compression ratio of the to-be-compressed block corresponding to the compressed code stream into a compressed code stream cache. In the embodiment of the invention, when decompressing the compressed code stream generated through compression, decompression is carried out according to the pixel points in the same to-be-compressed block, and thus, random access to the to-be-compressed blocks is realized.

Improved error resiliency of an encoding device, such as a video codec or encoder, operating in a compressed data transmission system, is achieved by enabling the encoding device to “shadow” or mimic the error conditions of a decoding device that receives and decodes compressed data sent by the encoding device. The encoding device is made aware of the specific error concealment scheme that the decoding device employs to reconstruct video frame data from a bit stream that contains bit or packet errors, and is also made aware of the location(s) in the frame to which the erroneous data corresponds and the particular manner in which the erroneous data was handled. The error-concealed data then can be used not only for the present encoding, but for subsequent encodings as well, to achieve synchronization between the encoder and decoder. Advantageously, the present invention improves on, but can work with, conventional error concealment schemes.

Methods and apparatus for artifact detection are provided by the present principles that measure the level of artifacts, such as those caused by temporal concealment of errors due to packet loss, for conditional error concealment. The principles are based on an assumption that sharp edges of video are rarely aligned with macroblock boundaries so video discontinuities are checked throughout the video. The scheme solves the problem of error propagation when temporal concealment of artifacts is used and the high false alarm rates of prior methods. Artifact detection methods are provided for regions of an image, an entire image, or for a video sequence, with error concealment provided conditionally based on the artifact levels.

The present invention relates to a method and device for providing encoded video data from a video signal. The method comprises the steps of providing intra-coded picture data and predictive-coded picture data, based on the video signal, and generating a first and a second frame of said encoded video data. In the generating step of the intra-coded picture data is arranged in first and second slices in the first and second frames, respectively. The slices are arranged in an overlapping manner in the frames, advantageously with a vertical overlap m_y which is equal to or greater than a maximum absolute length of a vertical motion vector. The invention also relates to a corresponding decoding method and device, as well as a corresponding video encoder, a video decoder and a video codec. The invention may be implemented and used in accordance with standard specifications such as H.264. The invention leads to increased network smoothness as well as improved robustness and reduced error propagation during transmission.

The invention relates to a decision feedback type equalizer, capable of better reducing error dissemination. The decision feedback type equalizer comprises a storage device, a logarithm likelihood ratio calculating part, a logarithm likelihood ratio evaluation part, a decrement calculating part, and a logarithm likelihood ratio attenuation part, wherein the storage device is used for keeping input signals that make the decision responding to multiple signs to be repeatedly operated; the logarithm likelihood ratio calculating part is used for calculating the logarithm likelihood ratio according to the signals obtained by deducting the output of the feedback filter from the input signals kept in the storage device; the logarithm likelihood ratio evaluation part, according to the absolute value of the logarithm likelihood ratio, carries out confirmation over the signals transmitting the bit that is evaluated with low reliability; the decrement calculating part calculates and keeps the logarithm likelihood ratio attenuation amount corresponding to the whole or part of bits of the signals affected by the feedback filter; and the logarithm likelihood ratio attenuation part makes the logarithm likelihood ratio carry out attenuation according to the logarithm likelihood ratio attenuation amount calculated in the previous decision feedback cycle.

The invention relates to the technical field of power systems, in particular to an enterprise distribution network power load forecasting network building method, a forecasting method and a system. The building method includes the steps: dividing distribution levels for a distribution network according to a power level structure of the distribution network and a power load distribution rule; setting data acquiring and computing terminals corresponding to power data acquiring nodes set in the distribution levels containing the data acquiring and computing terminals according to the distribution levels; building power load forecasting sub-networks by the data acquiring and computing terminals belonging to the same distribution level, and building a power load forecasting network by the power load forecasting sub-networks at all levels according to a preset network building rule. According to the enterprise distribution network power load forecasting network building method, the forecasting method and the system, when the power load of the enterprise distribution network is forecasted, power load forecasting accuracy can be improved, and fine forecasting effects can be achieved in the field of short-term load forecasting.

The invention provides a space-time coding method, a wireless signal transmitting, receiving and decoding method and a wireless signal transmitting, receiving and decoding device. The space-time coding method comprises the following steps: at a transmitting end, deserializing symbols in a serial code stream in a way of taking m symbols as a group to acquire m parallel code streams, wherein m is equal to (2*Nt-1), and Nt is greater than or equal to 2 and is the number of transmitting antennae; and forming an Nt*Nt coding matrix according to the m symbols, wherein the symbols in the m symbols belonging to a first parallel code stream are repeated for Nt times at positions of different columns and different rows of the code matrix, and the symbols in the m symbols belonging to 2ath and (2a+1)th parallel code streams are repeated for (Nt-a) times at the positions of the different columns and different rows of the code matrix respectively. The space-time coding method, and the wireless signal transmitting, receiving and decoding method and the wireless signal transmitting, receiving and decoding device can achieve higher spectrum efficiency and better transmission reliability.

A video fault tolerance coding method based on reversible information hiding and a redundancy technology is disclosed. The method comprises two steps: coding terminal information embedding and decoding terminal information extraction. The step of the coding terminal information embedding is characterized in that: redundant slice is adaptively added to odd numbered frame of a video sequence and the RS is not added to even numbered frame; a motion vector of the even numbered frame in a key area is extracted and converted into a bit sequence; a scheme for easily expanding bilateral reversible information hiding is provided, a motion vector distribution function in the RS is calculated and information of the even numbered frame is embedded in the motion vector of the RS in a previous frame. The step of the decoding terminal information extraction is characterized by: extracting the embedded information according to an extraction algorithm of the decoding terminal and a distribution function characteristic of the RS motion vector and converting into the motion vector which can be used by error concealment; according to a network state, if a packet loss phenomenon happens to the odd numbered frame, starting using the RS; if motion vector loss happens to the even numbered frame, using the extracted motion vector information to carry out error concealment. In the invention, the video fault tolerance coding solution scheme with a gradable character, a large capacity and a low bit rate expansion is provided.

According to the video coding and decoding system, method and device based on block recombination, the terminal and the medium provided by the invention; by changing the packaging sequence of the video data blocks, error propagation among the data blocks is reduced, and the damage of data packet damage and loss to video quality in network transmission can be reduced; under the condition of the same packet loss rate, objectively, the average peak signal-to-noise ratio value of a video picture restored by a receiving end is higher under most conditions; subjectively, the area of a defect region in a video picture is smaller, and large-area strip-shaped defects are converted into scattered small block defects. The method is obtained by adding new steps on the basis of an existing video coding and decoding algorithm, so that the method is wide in application range and is not limited to a certain specific video coding and decoding algorithm, and the video coding and decoding algorithm meeting the Internet video coding standard can be applied to improve the effect of the method. Due to the fact that the anti-error-code effect is achieved through reordering of the data blocks, redundant information is not added, and the network transmission capacity is saved.

The invention provides an auxiliary global positioning system (GPS) location method and system based on a wireless network single node. The method includes acquiring data of a mobile terminal at the current moment and a location result of the mobile terminal at the last moment; analyzing the acquired data, acquiring wireless location distance, the GPS position of the connected wireless node and the GPS location result, and calculating the wireless location result; searching for a route set in the wireless node covering range; conducting maximum likelihood estimation according to the GPS location result, the wireless location result and the route set to obtain the maximum likelihood position. An observation result is acquired through position prediction, wireless distance measuring and GPS location. The restraint condition is obtained through route searching. The method and system combines results of wireless location and GPS location based on the principle of the maximum likelihood estimation. Only one node is required for auxiliary location, the location errors can be reduced substantially, the location accuracy is improved, the wireless node number is effectively reduced, and the self-grouping wireless network location cost is reduced.

The invention particularly relates to a polarization code SCLF decoding method based on a flipping set. The invention belongs to the technical field of channel coding, and the method comprises the following steps: firstly, estimating LLR expectation of each polarization sub-channel and the bit error probability of an SC decoder by using a Gaussian approximation principle, then in actual decoding, obtaining the bit error probability of CA-SCL decoding through path measurement of the SCL decoder, and finally decoding the CA-SCL according to the bit error probability of the CA-SCL decoding. Error bits containing SC states are identified through the bit error probability, and a flipping set is obtained after the error bits are sorted according to an ascending order of reliability. And finally, when the CA-SCL decoding fails for the first time, flipping the bits in the flipping set in sequence by adopting a bit flipping scheme of only exchanging decision results on the SC state path, thereby reducing error propagation and decision errors of the CA-SCL decoding. Simulation results show that the SCLF decoding method based on the flipping set, provided by the invention, has the advantages that the decoding performance is obviously improved, and the calculation complexity is reduced.