Samsung Electronics Co., Ltd.

Maxim Koroteev

Elena Alshina

Alexander Alshin

Vadim Seregin

Nikolay Shlyakhov

Ekaterina Nesterova

A video encoding method and apparatus and a video decoding method and apparatus. In the video encoding method, a first predicted coding unit of a current coding unit that is to be encoded is produced, a second predicted coding unit is produced by changing a value of each pixel of the first predicted coding unit by using each pixel of the first predicted coding unit and at least one neighboring pixel of each pixel, and the difference between the current coding unit and the second predicted coding unit is encoded, thereby improving video prediction efficiency.

Image encoding method and device, and decoding method and device therefor

Image decoding device which obtains predicted value of coding unit using weighted average

Image decoding method which obtains a predicted value of a coding unit by weighted average of predicted values

Provided are a method and apparatus for correcting quantized coefficients. In the method, statistical values of coefficients and quantized coefficients are extracted from a received video data stream, coefficient correction values for each pixel position in blocks are determined by using the statistical distribution of the coefficients depending on the statistical values, and then the coefficients are corrected by respectively adding the coefficient correction values to corresponding coefficients of respective pixel positions.

Method and apparatus for correcting quantized coefficients in decoder

A method and apparatus for coding an image by using exponential Golomb binarization are provided. The method comprises: performing entropy coding on an input image by performing exponential Golomb binarization using a plurality of predetermined thresholds; measuring a bit rate of a bit stream which is generated by performing the entropy coding using each threshold; and determining an optimal threshold by comparing bit rates with regard to the thresholds.

Method and apparatus for encoding image by exponential Golomb binarization using optimal threshold, and method and apparatus for decoding image by exponential Golomb binarization using optimal threshold

A method of compressing image signals, to provide more efficient compression than a discrete cosine transform (DCT) while having the least amount of overhead information is provided. A partial exchange between a row and a column of a transform matrix is performed by applying several values to a parameter, which shows an extent of partial exchange between rows and columns of the transform matrix. When a value of the used angle parameter is stored or transmitted, an overhead increases. Processes generating a random sequence of the angle parameter, generating a compression coefficient matrix regarding each angle parameter of the generated random sequence, and calculating a compression rate are repeated to obtain an angle parameter with the highest compression rate. The angle parameter having the highest compression rate is not stored or transmitted, but a number in the random sequence of the corresponding angle parameter is obtained, and stored or transmitted.

Growing community of inventors

Suwon-si, South Korea

Maxim Koroteev