HOW TO CREATE NON OVERLAP SEGMENTS IN EMU8086 CODE
The method for transmitting data of claim 9, wherein the compressing the non-overlap segments using the Huffman coding algorithm comprises: defining symbols based on the residuals determining probability of symbol occurrences sorting the probabilities in a descending order determining a Huffman code tree building a Huffman codebook from the Huffman code tree updating Huffman code tree, and executing a codebook smoothing algorithm.ġ1. The method for transmitting data of claim 1, further comprising: compressing the non-overlap segments categorized as the unknown state using an adaptive entropy encoding algorithm, wherein the adaptive entropy encoding algorithm includes determining the smallest number of bits for each of the non-overlap segments using one of a Huffman coding algorithm, run length coding algorithm, a constant outleaver coding algorithm, and a dynamic range coding method.ġ0. The method for transmitting data of claim 1, further comprising: receiving an output of categorized segments from the state detection algorithm into a multi-state prediction model extracting multi-state prediction model parameters for each of the categorized segments generating predicted segments from the multi-state prediction model parameters and generating predicted residuals by subtracting the categorized segments from the predicted segments.ĩ. The method of claim 3, wherein the state detection algorithm categorizes one of the non-overlap segments as changing from a first constant state to a second constant states within the non-overlap segment.Ĩ. The method of claim 5, wherein the state detection algorithm categorizes one of the non-overlap segments as being in the periodical outlier state by: detecting number of outliers are less than a outlier threshold in the non-overlap segment, where outliers are identified when their values are beyond multiple times of standard deviation in the non-overlap segment, and after replacing the outliers using prediction values from their neighbor samples, detecting periodicity using the cross-correlation coefficients between the predefined minimum and maximum periods.ħ. The method of claim 1, wherein the state detection algorithm categorizes one of the non-overlap segments as being in the periodical state when the maximum cross-correlation coefficients between predefined minimum and maximum periods are beyond a periodicity threshold.Ħ. The method of claim 1, wherein the state detection algorithm categorizes one of the non-overlap segments as being in the constant outlier state when the number of outliers are less than a threshold in the non-overlap segment and the outliers are identified when their values are beyond multiple times of standard deviation in the non-overlap segment.ĥ. The method of claim 1, wherein the state detection algorithm categorizes one of the non-overlap segments as being in constant state when all data samples in the non-overlap segment have the same value.Ĥ. The method of claim 1, wherein compressing the one-dimensional data stream comprises: using a prediction model to generate predicted one-dimensional data determining a difference between the predicted one-dimensional data and the one-dimensional data and encoding the difference with a compression encoder.ģ.
A method for transmitting data, comprising: receiving data from two or more data sources selectively classifying the data into at least two input data streams, the at least two data streams including a one dimensional data stream, called an encoded data stream, and an imagery data stream, which can include still images or video or both and separately compressing the one-dimensional data stream into a first compressed bit-stream and the imagery data stream into a second compressed bit-stream combining the first compressed bit-stream and the second compressed bit-stream into a packetized bit-stream encrypting the packetized bit-stream to generate encrypted data packets prepared for transmission, wherein the one dimensional data stream is tiled into non-overlap segments using a state detection algorithm that categorizes the non-overlap segments into different states, wherein the different states include a constant state, a periodical state, a constant outlier state, a periodical outlier state, a step jump state, and an unknown state, and wherein corresponding encoders are used to compress each of the categorized segments.Ģ.
0 kommentar(er)
