Abstract:
화면간 예측 방법 및 장치가 개시된다. 상기 화면간 예측 방법은 예측 대상 블록에 대한 움직임 벡터 예측을 통해 구해진 PMV(Predicted Motion Vector)와, 상기 PMV를 기반으로 정화소(Integer Pixel) 단위의 움직임 예측을 통해 구해진 IMV(Integer Motion Vector) 간의 그라디언트(Gradient)를 계산하는 단계, 상기 그라디언트를 기반으로 기설정된 방향에 대해 부화소(Fractional Pixel) 단위의 보간을 통해 움직임 예측을 수행하는 단계 및 상기 움직임 예측을 기반으로 상기 예측 대상 블록의 최적 움직임 벡터를 도출하는 단계를 포함한다.
Abstract:
영상 부호화에서의 화면 내 예측 생략 방법이 개시된다. 상기 방법은 화면 내 예측을 수행할 현재 블록에 대한 주변 블록을 결정하는 단계 및 상기 주변 블록의 예측 모드 정보, 상기 주변 블록의 부호화 비용 정보, 및 상기 현재 블록과 상기 주변 블록 간의 경계에 위치한 경계 화소에 대한 연관성 정보 중 적어도 하나를 이용하여 상기 현재 블록에 대한 화면 내 예측을 생략할지 여부를 판단하는 단계를 포함한다.
Abstract:
Disclosed are a method and an apparatus for determining a coding mode. The method for determining a code mode includes the steps of: predicting a value of a distortion and an amount of bits generated when being coded in a current coding block (CB) in a candidate coding mode to estimate a rate-distortion (RD) cost; comparing the minimum RD cost produced based on the current CB with the estimated RD cost to determine whether a coding operation with respect to the candidate coding mode is omitted; and determining a final coding mode of the current coding block based on the minimum RD cost.
Abstract:
A bitstream analyzer is disclosed. The bitstream analyzer comprises: an analysis unit for receiving at least two encoded bitstreams, and analyzing encoding information and statistic information on each of at least two received encoded bitstreams; a comparison unit for comparing a difference between at least two encoded bitstreams based on each analysis result of at least two encoded bitstreams; and a GUI for outputting the analysis result by the analysis unit and the comparison result by the comparison unit, wherein the GUI outputs each analysis result of at least two encoded bitstreams in a split screen method at the same time, and outputs a difference between at least two encoded bitstreams as a graph.
Abstract:
The present invention provides a video encoding method using rate-distortion optimization. According to the video encoding method, a more accurate optimum encoding mode can be obtained by using the rate-distortion optimization with respect to an image output from an in-loop filter instead of using the rate-distortion optimization prior to the in-loop filter of a video encoding apparatus, and video encoding efficiency can be improved by reducing the bit rate required for reducing the difference between a finally decoded restoration image and an original image. [Reference numerals] (102) Intra-prediction unit; (104) Motion compensation unit; (106) Motion prediction unit; (110) Transformation unit; (112) Quantization unit; (114) Inverse quantization unit; (116) Inverse transformation unit; (120) In-loop filter; (122) Reference image buffer; (124) Entropy encoding unit; (AA) Input image; (BB) Intra; (CC) Inter; (DD) Output bit stream
Abstract:
PURPOSE: An image encoding device for high speed prediction and a device thereof are provided to improve a processing speed of intra prediction. CONSTITUTION: An intra prediction unit calculates each Sum of Absolute Transform Difference (SATD)-based encoding cost of a plurality of intra prediction modes for a current prediction unit. The intra prediction unit determines one or more first candidate encoding modes for the current prediction unit based on the SATD-based encoding cost (S310). The intra prediction unit determines a final encoding mode among the first candidate encoding modes (S320). [Reference numerals] (S310) Determine a first candidate encoding mode; (S320) Determine a final encoding mode; (S330) Determine the size of a conversion unit; (S340) Encode
Abstract:
PURPOSE: An interlayer intra prediction method and an apparatus thereof perform interlayer intra prediction by using intra prediction data of a periphery block, thereby improving encoding efficiency. CONSTITUTION: A layer determining unit (910) determines a reference layer for interlayer intra prediction. The layer determining unit selects a candidate layer of the highest layer as the reference layer. A prediction data generator (920) produces intra prediction data from the reference layer. The prediction data generator produces intra prediction data based on an intra-predicted co-located block among front-to-back frames of the reference frame. An intra prediction part (930) performs interlayer prediction about a current block based on the intra prediction data. [Reference numerals] (910) Layer determining unit; (920) Prediction data generator; (930) Intra prediction part
Abstract:
PURPOSE: A method and device for detecting an advertisement image are provided to detect the advertisement image at high speed and reduce the size of feature information and the sensitivity about the change of a pixel value. CONSTITUTION: An advertisement image detecting apparatus divides a plurality of frames of a monitoring target image into partial blocks with a preset size(120). The advertisement image detecting apparatus produces feature information about the partial blocks(130). The advertisement image detecting apparatus converts the feature information to feature information vectors(140). The advertisement image detecting apparatus detects reference advertisement image by comparing the feature information vector of the reference advertisement image with the above feature information(150).