50results about How to "Reduce the theoretical limit entropy" patented technology

The invention relates to a method for adaptive texture gradient prediction in bandwidth compression, which includes selecting N sampling methods to sample pixel components in the current MB; selecting M prediction methods to predict the current MB; and calculating the current MB respectively. The prediction residual and SAD; select the sampling mode and prediction mode of the current MB according to the SAD. The present invention calculates the prediction residual and SAD of the current prediction macroblock by defining the sampling mode of the MB. Compared with the existing methods, when the texture of the image to be compressed is more complex, for the MB at the texture boundary of the current image, according to the texture gradient principle, it does not depend on the surrounding MB of the current MB, but through the current MB The prediction residual obtained by its own texture characteristics can improve the accuracy of the prediction residual value for complex texture areas, further reduce the theoretical limit entropy, and increase the bandwidth compression rate.

The invention relates to a dual-mode selection prediction method for a complex texture in bandwidth compression, which comprises the steps of: partitioning a to-be-encoded video image into a pluralityof macro blocks, and determining a to-be-encoded pixel component; by adopting an adaptive template prediction method, selecting a first reference pixel of each currently encoded pixel in a currentlyencoded macro block in an adaptive template, and calculating to obtain a group of first predicted residuals; by adopting an adaptive rectangular window prediction method, selecting a second referencepixel of each currently encoded pixel in the currently encoded macro block in a rectangular prediction search window, and calculating to obtain one group of second predicted residuals; calculating a first subjective difference according to one group of first predicted residuals, and calculating a second subjective difference according to one group of second predicted residuals; and comparing the first subjective difference with the second subjective difference, and determining an optimal prediction method of the currently encoded macro block to obtain one group of optimal predicted residuals.The dual-mode selection prediction method uses the macro blocks as prediction units, and according to different texture characteristics in different regions of the image, the optimal prediction methods are adaptively selected, so that the dual-mode selection prediction method is better in prediction effect.

The present invention relates to a method for predicting adaptive texture gradients in bandwidth compression, comprising: selecting N sampling methods to sample the current MB, wherein the value of N is a natural number greater than 1; predicting the current MB, and obtaining the The prediction residual of the current MB; calculating the sum of the absolute value of the residual of the current MB; and determining the sampling mode of the current MB according to the sum of the absolute value of the residual. The present invention adopts multiple sampling methods for the current MB to obtain the prediction residual and SAD of the current MB. When processing a compressed image with a complex texture, for the current MB at the texture boundary of the current image to be compressed, due to the differences between the current MB and the surrounding MB Not in the same texture area, resulting in a poor correlation between the current MB and surrounding MBs, but the present invention does not depend on the MBs around the current MB based on the texture gradient principle, but obtains the prediction residual through the texture characteristics of the current MB itself, It can improve the accuracy of prediction residual value for complex texture area.

The invention relates to a prediction method based on video compression, comprising: obtaining the first residual absolute value sum of the to-be-predicted MB; obtaining the second residual absolute value sum of the to-be-predicted MB; obtaining the first standard deviation and the second standard deviation of the to-be-predicted MB; obtaining a first residual subjective sum according to the firstresidual absolute value and the first standard deviation; obtaining a second residual subjective sum according to the second residual absolute value and the second standard deviation; selecting a minimum value of the first residual subjective sum and the second residual subjective sum; and using the prediction method corresponding to the minimum value to predict the to-be-predicted MB. Accordingto the prediction method based on video compression, a variety of prediction methods are selected for predicting the to-be-compressed image, so that an optimal prediction method can be selected from avariety of prediction methods when the texture of the to-be-compressed image is complex, and therefore, the portion of the to-be-compressed image where the texture is complex can be predicted, and the compression quality of the to-be-compressed image can be guaranteed.

The invention relates to a prediction method based on video compression, which comprises the steps of obtaining first standard deviation and second standard deviation of an MB to be predicted; obtaining the first sum of absolute difference of the MB to be predicted by using first prediction residuals of the MB to be predicted; obtaining the second sum of absolute difference of the MB to be predicted by using second prediction residuals of the MB to be predicted; obtaining the first subjective difference according to the first standard deviation and the first sum of absolute difference; obtaining the second subjective difference according to the second standard deviation and the second sum of absolute difference; and determining a final prediction method for the MB to be predicted accordingto the first subjective difference and the second subjective difference. Various prediction methods are adopted for the same image to be compressed, and the image to be compressed adaptively selectsone prediction method for prediction according to the texture correlation, thereby enabling the image to be compressed to select a prediction method for prediction in a targeted manner, and greatly improving the quality of video compression.