Signal transforming method and device

Abstract

Provided are a signal transforming method and a signal transforming device. For example, the signal transforming method includes determining a minimum-value matrix and a maximum-value matrix with respect to elements of a matrix used in frequency transformation, wherein the minimum-value matrix is configured of elements of minimum value and the maximum-value matrix is configured of elements of maximum value; determining a maximum threshold value of a result value of a function indicating at least one selected from transform distortion, normalization, and orthogonality of the matrix; determining a transform matrix configured of elements that are greater than the elements of the minimum-value matrix and less than the elements of the maximum-value matrix at respective positions of the matrix, and in which the result value of the function is less than the maximum threshold value; and transforming an input signal by using the determined transform matrix.

1 . A signal transforming method comprising:
 determining a minimum-value matrix and a maximum-value matrix with respect to elements of a matrix used in frequency transformation, wherein the minimum-value matrix is configured of elements of minimum value and the maximum-value matrix is configured of elements of maximum value;   determining a maximum threshold value of a result value of a function indicating at least one selected from transform distortion, normalization, and orthogonality of the matrix;   determining a transform matrix configured of elements that are greater than the elements of the minimum-value matrix and less than the elements of the maximum-value matrix at respective positions of the matrix, and in which the result value of the function is less than the maximum threshold value; and   transforming an input signal by using the determined transform matrix.   
     
     
         2 . The signal transforming method of  claim 1 , wherein the determining of the minimum-value matrix and the maximum-value matrix comprises:
 determining, based on a size of the matrix, a DCT matrix that is a matrix used in a DCT transform; and   determining, by using the determined DCT matrix, the minimum-value matrix and the maximum-value matrix which are configured of integer elements.   
     
     
         3 . The signal transforming method of  claim 2 , wherein the determining, by using the determined DCT matrix, of the minimum-value matrix and the maximum-value matrix comprises:
 determining the minimum-value matrix as a matrix comprising element values obtained by multiplying each of elements of the determined DCT matrix by a predetermined factor, rounding multiplication results of the multiplying, and then subtracting a predetermined value from results of the rounding; and   determining the maximum-value matrix as a matrix comprising element values obtained by multiplying each of the elements of the determined DCT matrix by the predetermined factor, rounding the results of the multiplying, and then adding the predetermined value to the results of the rounding.   
     
     
         4 . The signal transforming method of  claim 3 , wherein the determining of the maximum threshold value comprises:
 determining a reference matrix comprising the element values obtained by multiplying each of the elements of the determined DCT matrix by the predetermined factor, and rounding the results of the multiplying;   determining a reference matrix function value that is a result value of a function indicating at least one selected from transform distortion, normalization, and orthogonality of the reference matrix; and   determining the reference matrix function value as the maximum threshold value.   
     
     
         5 . The signal transforming method of  claim 1 , wherein the determining of the maximum threshold value comprises:
 determining a reference matrix that is a matrix configured of integer elements, based on a method defined in H.265;   determining a reference matrix function value that is a result value of a function indicating at least one selected from transform distortion, normalization, and orthogonality of the reference matrix; and   determining the reference matrix function value as the maximum threshold value.   
     
     
         6 . The signal transforming method of  claim 1 , wherein the determining of the maximum threshold value comprises:
 determining the maximum threshold value by multiplying a number of rows of the transform matrix by a predetermined value.   
     
     
         7 . The signal transforming method of  claim 1 , wherein the transforming of the input signal by using the determined transform matrix comprises:
 determining an input matrix that is a matrix with respect to the input signal; and   performing an operation on the determined transform matrix and the input matrix, and thus determining an output matrix that is a matrix with respect to an output signal.   
     
     
         8 . The signal transforming method of  claim 1 , wherein the transform matrix is a matrix configured of {{32, 32, 32, 32}, {42, 17, −17, −42}, {32, −32, −32, 32}, {17, −42, 42, −17}}. 
     
     
         9 . A signal transforming device comprising:
 a range determiner configured to determine a minimum-value matrix and a maximum-value matrix with respect to elements of a matrix used in frequency transformation, wherein the minimum-value matrix is configured of elements of minimum value and the maximum-value matrix is configured of elements of maximum value;   a maximum threshold value determiner configured to determine a maximum threshold value of a result value of a function indicating at least one selected from transform distortion, normalization, and orthogonality of the matrix;   a transform matrix determiner configured to determine a transform matrix configured of elements that are greater than the elements of the minimum-value matrix and less than the elements of the maximum-value matrix at respective positions of the matrix, and in which the result value of the function is less than the maximum threshold value; and   a transformer configured to transform an input signal by using the determined transform matrix.   
     
     
         10 . The signal transforming device of  claim 9 , wherein the range determiner is further configured to determine, based on a size of the matrix, a DCT matrix that is a matrix used in a DCT transform, and determine, by using the determined DCT matrix, the minimum-value matrix and the maximum-value matrix which are configured of integer elements. 
     
     
         11 . The signal transforming device of  claim 10 , wherein the range determiner is further configured to determine the minimum-value matrix as a matrix comprising element values obtained by multiplying each of elements of the determined DCT matrix by a predetermined factor, rounding multiplication results of the multiplying, and then subtracting a predetermined value from results of the rounding, and
 determine the maximum-value matrix as a matrix comprising element values obtained by multiplying each of the elements of the determined DCT matrix by the predetermined factor, rounding the results of the multiplying, and then adding the predetermined value to the results of the rounding.   
     
     
         12 . The signal transforming device of  claim 11 , wherein the maximum threshold value determiner is further configured to determine a reference matrix comprising the element values obtained by multiplying each of the elements of the determined DCT matrix by the predetermined factor, and rounding the results of the multiplying,
 determine a reference matrix function value that is a result value of a function indicating at least one selected from transform distortion, normalization, and orthogonality of the reference matrix, and   determine the reference matrix function value as the maximum threshold value.   
     
     
         13 . The signal transforming device of  claim 9 , wherein the maximum threshold value determiner is further configured to determine a reference matrix that is a matrix configured of integer elements, based on a method defined in H.265,
 determine a reference matrix function value that is a result value of a function indicating at least one selected from transform distortion, normalization, and orthogonality of the reference matrix, and   determine the reference matrix function value as the maximum threshold value.   
     
     
         14 . The signal transforming device of  claim 9 , wherein the maximum threshold value determiner is further configured to determine the maximum threshold value by multiplying a number of rows of the transform matrix by a predetermined value. 
     
     
         15 . The signal transforming device of  claim 9 , wherein the transformer is further configured to determine an input matrix that is a matrix with respect to the input signal,
 determine a matrix transposed from the determined transform matrix, and   perform an operation on the matrix transposed from the determined transform matrix and the input matrix, and thus determine an output matrix that is a matrix with respect to the output signal.