54 results about "Inverse discrete cosine transformation" patented technology

This invention relates to methods of feature extraction from MPEG-2 and MPEG-4 compressed video sequences. The spatio-temporal compression complexity of video sequences is evaluated for feature extraction by inspecting the compressed bitstream and the complexity is used as a descriptor of the spatio-temporal characteristics of the video sequence. The spatio-temporal compression complexity measure is used as a matching criterion and can also be used for absolute indexing. Feature extraction can be accomplished in conjunction with scene change detection techniques and the combination has reasonable accuracy and the advantage of high simplicity since it is based on entropy decoding of signals in compressed form and does not require computationally expensive inverse Discrete Cosine Transformation (DCT).

A method of downscale decoding MPEG-2 video includes an Inverse Discrete Cosine Transformation (DCT) procedure for performing a ½ horizontal downscaling to convert DCT coefficients in a 8×8 array block of the video into a 4×8 array intra-block and performing a ½ vertical downscaling to convert the intra-block into an intra-coded picture having ¼ resolution of the original; and a downscaling motion compensation procedure for performing a motion compensation to the current intra-block to obtain a predictive block having ½ horizontal size of the original, adding the predictive block with a residual block produced by the same method applied to the intra-blocks to obtain a 4×8 array inter-block, and performing a ½ vertical downscaling to the inter-block for outputting a predictive-coded picture and a bidirectional predictive-coded picture having ¼ resolution of the original, so as to simplify the complexity of the decoding computation and enhance the decoding speed.

The invention provides an MPEG2 (Moving Picture Experts Group 2) video watermarking realization method based on a DC (Discrete Cosine) coefficient, belonging to the field of computer information security. According to the method, watermarking information is embedded into a luminance component of an I frame of an encoded video so as to ensure the robustness of a video watermarking; the last difference DC coefficient of a Y component of the last macro block of each strip which has the smallest influence on the video quality is selected at the embedding position of the watermarking, so that video data of other strips is not influenced after the DC coefficient is changed; in an embedding and extracting algorithm of the watermarking, watermarking information is embedded on the basis that encoding quantity of a DCT (Discrete Cosine Transformation) coefficient is not changed before and after the watermarking is added by combining the DC coefficient in an MPEG2 video and utilizing a Huffman difference encoding characteristic; the embedding and the extraction of the watermarking do not relate to lots of data DCT and IDCT (Inverse Discrete Cosine Transformation) conversion, quantization and inverse quantization, the complexity of a video watermarking system is reduced greatly, the embedding and extracting efficiency of the watermarking is improved and the accuracy of watermarking information extraction is enhanced.

The invention provides a network image copyright real-time distinguishing method. Copyright protection is necessary for digital images during network communication, but the existing robust digital watermarking technology can not achieve the real-time copyright distinguishing. The method comprises the following steps: partitioning an original image to sub-blocks which are not overlapped with each other at an embedded end; processing each sub-block by discrete cosine transformation; generating a feature watermark by comparing the DC coefficient of each sub-block with the average DC coefficient of all sub-blocks; encrypting the Logistic chaotic sequence of the feature watermark; adjusting the magnitudes of two low-frequency coefficients of the discrete cosine transformation of each sub-blockand self-embedding the encrypted feature watermark; and processing each sub-block by inverse discrete cosine transformation to obtain images with watermarks. The method has high robustness to attack.Substantially, the method provided by the invention can achieve complete blind detection by combining the self-embedding encrypted feature watermark with watermark blind extraction and authentication, thereby achieving real-time copyright distinguishing of network images.

A method and apparatus for processing a gray image, including: calculating a mean value and a standard difference of each edge pixel needing processing in an original gray image, and determining a predefined noise variance of each edge pixel according to the mean value and the standard difference; performing two-dimensional discrete cosine transform on an original gray image of a first adjacent area of each edge pixel; performing Wiener filtering on the two-dimensional discrete cosine transformed image data according to the noise variance; performing discrete cosine inverse transform on the two-dimensional Wiener filtered image data to obtain a filtered gray image of the first adjacent area; and extracting a filtered pixel value of each edge pixel from the filtered gray image of the first adjacent area, and obtaining a processed pixel value of each edge pixel by performing weighted summation on the filtered pixel value and an original pixel value.

The invention relates to a video softcast method based on a channel characteristic. According to the method, subcarriers of the highest quality of a video data allocation channel are transmitted in a physical layer, and a decoding terminal uses a linear least square estimator (LLSE) to denoise and uses zero forcing decision feedback equalization (ZF-DFE) to decode, so that the video recovery quality is improved finally. The video softcast method comprises following steps that a coding terminal reads in a video sequence and partitions images, removes intra-frame redundancy through discrete cosine transformation (DCT), performs power distribution so as to minimize reconstruction errors, and finally generates data of the same significance through hadamard transform; acquired data packages are sent into orthogonal frequency division multiplexing (OFDM), data of each package is simply modulated firstly so as to form a series of symbols, subcarriers are allocated according to CSI transmitted by a feedback channel, serial-parallel conversion, inverse fast fourier transform (IFFT), cyclic prefix insertion and parallel-serial conversion are performed on the symbols so as to obtain OFDM transmission symbols, and the transmission symbols are transmitted into additive white Gaussian noise (AWGN) to be sent to a receiving terminal; the process of the receiving terminal is opposite to the process of a transmitting terminal; and the decoding terminal uses the LLSE to denoise, and uses ZF-DFE to decode, and inverse discrete cosine transformation (IDCT) is performed finally, so as to acquire the recovered video.

The invention relates to the field of development of a video multimedia technology, and discloses an inverse discrete cosine transformation (IDCT) method and apparatus. In the method, through counting the number of nonzero coefficients before IDCT, according to the number of the nonzero coefficients, IDCT processing can be carried out on variable coefficients, and different IDCT processes rather than full-coefficient IDCT processing are employed for discrete cosine transformation (DCT) nonzero coefficients of different blocks, so that the IDCT processing speed can be improved, and the performance of a multimedia decoder is enhanced.

The invention discloses an information hiding method and apparatus, a detection method and a device, and an anti-counterfeiting tracing method. The information hiding method comprises the steps of converting hidden information into a hidden information binary image, and performing coding scrambling; discrete cosine transform is performed on the image sub-blocks in the blue channel image of the target area image according to the hidden information binary image; modifying the intermediate frequency coefficients meeting preset requirements according to pixel values of pixels in the hidden information binary image corresponding to the image sub-blocks to obtain a modified coefficient matrix; and performing inverse discrete cosine transform on the modified coefficient matrix to obtain a blue channel image of the target area image embedded with the hidden information, and replacing the blue channel image of the target area image to obtain a carrier image embedded with the hidden information.According to the method, the invisibility of the hidden information of the transform domain is combined, the intermediate frequency coefficients meeting the preset requirements in the image sub-blocks are modified, and the hidden information is embedded into the carrier image, so that the robustness of information hiding is improved.

The invention discloses an RGB565 true color image lossy compression and decompression method. Firstly, original RGB image data are separated into monochromatic R ordered data, monochromatic G ordered data and monochromatic B ordered data and are arranged into an 8*8 matrix, then discrete cosine transformation is performed on the data, rounding is directly performed on direct-current components generated after transformation, data quantization is performed on alternating-current components, Z-shaped scanning is performed on quantized matrix data blocks, and zero values in the data are integrated; secondly, compression coding is performed on the quantized data, and Z-shaped inverse scanning is performed after decoding is performed; finally, inverse discrete cosine transformation is performed on the data, the data components are combined, and decompressed RGB image data are obtained. Under the premise that it is guaranteed that images are undistorted, the method increases the compression ratio of the image data, and can obtain clear real characters in a decompressed mode, reduce transmission amount of the data, and save transmission time of the data.

The invention relates to a multiplying unit structure for a discrete cosine transformation (DCT)/inverse discrete cosine transformation (IDCT) circuit under a high efficiency video coding (HEVC) standard. The multiplying unit structure comprises a preprocessing operation module, a primary shifting adding operation module and a secondary shifting adding operation module which are connected in series in sequence. A pre-operation unit is provided with a data input port and a parameter input port. The secondary shifting adding operation module is provided with an operation result output port. Compared with the prior art, the multiplying unit structure has the advantages of being capable of meeting system functions, having smaller occupancy of hardware resource and faster operation cycle, and the like.

A high-speed inverse discrete cosine transformation method and apparatus are provided. All elements of a discrete cosine transformation (DCT) matrix for elements having a value other than 0 are searched for in a predetermined order when a total number of elements having a value other than 0 is not greater than a predetermined critical value. Two-dimensional (2D) IDCT is performed on the elements having a value other than 0. 2D IDCT is performed on the DCT matrix when the total number of elements having a value other than 0 is greater than the predetermined critical value.

The present invention provides for an improved video compression and encoding that optimizes and enhances the overall speed and efficiency of processing video data. In one embodiment, the video codec transmits the output of an entropy decoder to a lossless compressor and memory before going through inverse discrete cosine transformation and motion compensation blocks.

An image processing apparatus comprises a central processing unit (CPU) and a graphics processing unit. The CPU comprises a first table specification, a second table specification, an inverse discrete cosine transformation (IDCT) module, a color space converting module and an entropy decoder. The entropy decoder retrieves and processes input image data as a two-dimensional quantized image according to the first table specification. The graphics processing unit performs a dequantizing operation on the input image data according to the second table specification. The IDCT module performs an IDCT operation on the dequantized input image data. The color space converting module performs a color space conversion on the input image data and outputs the converted image data.

An image compression and expansion apparatus reproduces an expanded image corresponding to an original image from a reduced image. The original image data is comprised of approximately 1000×600 pixel values Pyx. A first matrix M1 comprised of 64×64 pixel values Pyx is extracted from the original image data. Regarding the original image data arranged in the first matrix M1, an average value of 8×8 pixel values Pyx forming a block B1 is obtained, and thus a second matrix M2 is generated using the average value as a single pixel value. The pixel values of the second matrix M2 is subject to two dimensional discrete cosine transformation to obtain a matrix MD comprised of 8×8 DCT coefficients. Expanded two dimensional inverse discrete cosine transformation is applied to the matrix MD to obtain a third matrix M3 comprised of 64×64 pixel values. The expanded image data comprising the third matrix M3 is transformed to the same coordinate system as the original image data and recorded on to the recording medium.

The invention discloses a channel estimation method, a device and a related detection system. The channel estimation method comprises the following steps: obtaining the channel parameter of a pilot frequency symbol in a DCT (discrete cosine transformation) domain; filtering the channel parameter in the DCT domain; carrying out IDCT (inverse discrete cosine transformation) on the filtered channel parameter; according to the decay coefficient of the channel frequency domain on a pilot frequency symbol position, obtaining the decay coefficient of the channel frequency domain on a data symbol. The channel estimation device comprises a module 1, a module 2, a module 3 and a module 4, wherein the module 1 is used for obtaining the channel parameter for the DCT domain of the pilot frequency symbol; the module 2 is used for filtering the channel parameter in the DCT domain; the module 3 is used for carrying out IDCT on the filtered channel parameter; and the module 4 is used for obtaining the channel parameter on the data sign position according to the channel parameter on the pilot frequency symbol position. The related detection system comprises the channel estimation device. In the embodiment of the invention, because DCT variation has the action of energy concentration, the channel parameter energy performed DCT variation is centralized in the front half part of a vector so as to change flat bottom effect to improve channel estimation performance.

The invention discloses a volume data multi-watermark method based on three-dimensional DWT-DCT (3D Wavelet Transform-Discrete Cosine Transformation) perceptual hashing, and belongs to the field of multimedia signal processing. The method comprises the following steps: firstly, performing 3DDWT-DCT on medial volume data, selecting previous 4*4*4 coefficients, further performing 3D-IDCT (Three Dimensional-Inverse Discrete Cosine Transformation), subsequently extracting a perceptual hashing value with a robust characteristic in an inverse transformation coefficient through perceptual hashing, associating a multi-watermark sequence with the perceptual hashing value to obtain a string of two-value key sequences, subsequently storing the two-value key sequences in a third party, extracting the three-dimensional DWT-DCT perceptual hashing value from volume data to be tested, and associating the extracted perceptual hashing value with the two-value key sequences stored in the third party so as to extract multiple watermarks. The method is a volume data multi-digital watermarking technique based on three-dimensional DWT-DCT perceptual hashing, and has good robust property, and the content of original volume data is not changed by embedding of multiple watermarks.

The invention relates to an automatic environmental identification method used for a hearing aid. According to the invention, a signal accessed by a microphone of a hearing aid is sampled according to a preset sampling frequency so as to form continuous multiframe data; and then after each frame of a multiframe signal passes through a Hamming window of a preset point, DFT is carried out on the processed multiframe data; processing including a triangle overlapping window and a triangle MEL weighted filter set of a preset class number is carried out on an energy spectrum of data that has been processed by DFT so as to obtain a preset number of MEL frequency spectrum points; processes of taking a logarithm and inverse discrete cosine transformation are carried out on the preset number of MEL frequency spectrum points so as to obtain a hidden Markov model (HMM) coefficient; at last, a highest probability of a single path is recorded through a Viterbi algorithm by taking the obtained HMM coefficient as a characteristic and employing an HMM containing an initial state and an end state and three radiating states, so that an optimum state sequence is found and thus an environment which the signal accessed by the microphone are from is determined.

An MPEG video decoding method and an MPEG video decoder are provided. The method includes determining whether to perform motion compensation on motion-vector-decoded data or not depending on a value of a decoded motion vector, determining whether to perform inverse discrete cosine transformation (IDCT) on motion-compensated data or not depending on values of decoded DCT coefficients, and generating a decoded image based on the results of the two determinations.

The invention discloses invisible watermark image construction and classification methods, invisible watermark backdoor attack model construction and classification methods and a system, and the invisible watermark image construction method comprises: carrying out the Haar discrete wavelet transformation of an original image, and obtaining a low-frequency information matrix, a horizontal high-frequency information matrix, a vertical high-frequency information matrix and a diagonal high-frequency information matrix of the original image; performing block discrete cosine change on the low-frequency information matrix to obtain a plurality of low-frequency information matrixes; and encrypting a watermark image to obtain an encrypted watermark image, embedding the encrypted watermark image into a plurality of low-frequency information matrixes by adopting a singular value decomposition watermark algorithm, and then obtaining an invisible watermark image through block discrete cosine inverse transformation and Haar discrete inverse wavelet transformation. Based on an invisible watermark technology, a more hidden backdoor is generated by utilizing the weakness that the neural network is easily attacked by the backdoor, so that the classification accuracy of the neural network is reduced to a certain extent, and the invention has great significance in the field of artificial intelligence security.

The invention discloses a sound wave information decoding method and device, and the method comprises the steps: carrying out the real-time decompression of a to-be-decoded audio quantized compresseddata stream, so as to generate one or more continuous quantized data blocks Zx, respectively selecting n quantized numerical values with the same frequency or the most approximate frequency corresponding to the n bit signal frequencies from each quantized data block Zx, and forming quantized data by the obtained n*x quantized numerical values, carrying out quantized reduction on each piece of quantized data to obtain corresponding energy data, carrying out inverse discrete cosine transform processing on each piece of energy data to obtain amplitude data, and carrying out sound wave signal decoding on the amplitude data to obtain corresponding sound wave data. By utilizing the method and the device, the operation speed of performing acoustic decoding on the audio quantized compressed data stream by the interpretive language can be remarkably improved.