JP4079842B2 - Data multiplexing / extraction method for information data with respect to enlarged image data and apparatus to which these are applied - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a method and apparatus for enlarging original image data, multiplexing other data on the original image data, and extracting other data from the multiplexed data.
[0002]
[Prior art]
Currently, with the progress of digital broadcasting, high-definition high-definition TV is becoming common in TV broadcasting stations. Up-conversion (hereinafter abbreviated as up-conversion) of conventional SDTV images is expanded and converted to HDTV format. There is a lot to do. When up-conversion, the problem is how to re-multiplex data such as subtitle broadcast and text broadcast information multiplexed during the video blanking period of the SDTV signal into the HDTV signal. At that time, it is conceivable to re-multiplex in the blanking period of HDTV, but conversion is necessary because the standard is different.
[0003]
As one of the methods for solving the problem, a method of multiplexing the up-conversion image can be considered. In a conventional image data multiplex processing method (Japanese Patent Laid-Open No. 2001-78008), predetermined image data or LSB (least significant bit) of all image data is replaced with binary information data. An example in which this method is applied to an enlarged image is shown in FIG. First, after upsampling the original image, the image is enlarged by performing an interpolation filter process. A conventional method is applied to the enlarged image, and information data is multiplexed on the LSB.
[0004]
[Patent Document 1]
Japanese Patent Laid-Open No. 2001-78008
[Problems to be solved by the invention]
However, in the conventional method, since the LSB is replaced with the information data, this corresponds to reducing the number of bits representing the pixel value by 1 bit, and the image quality of the data multiplexed image is deteriorated. In addition, it is impossible to restore the image quality after extracting the information data. For this reason, there is a problem that the high image quality, which is one of the merits of digital broadcasting, is impaired by multiplexing of information data.
[0006]
That is, in the prior art, no consideration is given to image quality degradation of the image. Even if the pixels on which data is multiplexed are limited, the error from the true value is 1 or more at the maximum in the data multiplexed pixels. In addition, it is impossible to restore the image quality after extracting the information data.
[0007]
SUMMARY OF THE INVENTION Accordingly, an object of the present invention is to provide a method and apparatus for multiplexing information data while reducing image degradation, which is the above-mentioned problem, and an extraction method for extracting multiplexed information data and restoring the image quality. To provide an apparatus.
[0008]
[Means for Solving the Problems]
According to the first aspect of the present invention, the interpolation data for the adjacent image data is obtained from the adjacent original image data based on the original image data, and when the interpolation data includes a value after the decimal point, the value after the decimal point is discarded. Information data is added to the least significant bit of the interpolation data only for interpolation data that has been subjected to decimal point processing and rounded down to the nearest decimal point. It is possible to obtain a data multiplexing method for information data with respect to enlarged image data, which is characterized by being inserted between original image data.
[0009]
According to the second aspect of the present invention, the multiplexed image data multiplexed by the data multiplexing method is received, original image data is generated by downsampling, and reproduction interpolation data is obtained from adjacent original image data based on the original image data. And interpolating data from the received multiple image data, and when the reproduction interpolation data includes a value after the decimal point, comparing the least significant bit in the integer part of the reproduction interpolation data and the reproduction interpolation data, A data extraction method for information data with respect to an enlarged image is obtained, wherein a logical value of 0 level is output as information data when they match, and a logical value of 1 level is output as information data when they do not match. be able to.
[0010]
According to invention of Claim 3, the program for making a computer perform the data multiplexing method of the information data with respect to the expansion image data described in Claim 1 can be obtained.
[0011]
According to invention of Claim 4, the program for making a computer perform the data extraction method of the information data with respect to the expansion image data described in Claim 2 can be obtained.
[0012]
According to the fifth aspect of the present invention, an upsampling unit that obtains interpolation data for the adjacent image data from the adjacent original image data based on the original image data, and when the interpolation data includes a value after the decimal point, Only for the interpolation data obtained by rounding down the value of the interpolation data and the interpolation data obtained by rounding down the value after the decimal point, the information data is added to the least significant bit of the interpolation data, and the interpolation data after the decimal point processing or the decimal point processing is not performed. It is possible to obtain a data multiplexing apparatus for information data with respect to enlarged image data, characterized in that each of the interpolation data is inserted between the adjacent original image data.
[0013]
According to the sixth aspect of the present invention, the down-sampling unit that receives the multiplexed image data multiplexed by the data multiplexing device and generates the original image data by down-sampling, and the adjacent original image data based on the original image data An upsampling unit that generates reproduction interpolation data from the image, an interpolation data extraction unit that extracts interpolation data from the received multiplexed image data, and a minority presence / absence determination that determines whether the reproduction interpolation data includes a value after the decimal point And when the reproduction interpolation data is determined to contain a value after the decimal point by the determination, the interpolation data and the least significant bit in the integer part of the reproduction interpolation data are compared. A data integer that outputs a logical value of 0 level and outputs a logical value of 1 level as information data in case of mismatch And the LSB comparator unit, it is possible to obtain a data extraction device of the information data for the enlarged image data, characterized in that it comprises a.
[0014]
According to a seventh aspect of the present invention, there is provided a data extraction device for information data for enlarged image data according to the sixth aspect, wherein the comparison in the data integer part LSB comparison unit is performed by an exclusive OR circuit. Can do.
[0015]
DETAILED DESCRIPTION OF THE INVENTION
DESCRIPTION OF EXEMPLARY EMBODIMENTS Hereinafter, exemplary embodiments of the invention will be described with reference to the drawings. However, the dimensions, materials, relative arrangements, and the like of the components described in this embodiment are not intended to limit the scope of the present invention only to those unless otherwise specified.
[0016]
FIG. 1 is a block diagram of an apparatus for multiplexing character data for an original image according to an embodiment of the present invention. The data multiplexing apparatus 1 receives two systems of original image data and information data, and inserts information data into the original image data by a process described later. In the system of original image data, an upsampling unit 3 that upsamples the original image data, a filter 4 that interpolates an intermediate value of adjacent original image data using this output as input, and an interpolation value that includes a fractional part Includes a minority processing unit 5 that performs a truncation process. On the other hand, the information data is received by the information data unit 7, and the information data is output in accordance with the decimal point truncation process. Then, according to the processing of the minority processing unit 5, a data multiplexing unit 6 that multiplexes the information data from the information data unit 7 and the original image data including the interpolation value is provided.
[0017]
Next, the operation of the data multiplexing apparatus 1 will be described using the flowchart shown in FIG. In the following, it is assumed that a digital image expressed in 8 bits is input as original image data, and the original image data is expanded in the horizontal direction or the vertical direction, and information when generating double image data Data multiplexing will be described. That is, at the time of enlargement, one interpolation data is generated for the adjacent original image data (2 pixels) to be input. Here, as an interpolation method for storing pixel values, linear interpolation is adopted, and it goes without saying that various interpolation methods are applied by known techniques. This is because the essence of the present invention is not a variety of interpolation methods, but is characterized by multiplexing of information data so that image quality does not deteriorate after interpolation data is obtained by interpolation processing.
[0018]
The upsampling unit 3 upsamples the received original image data, and generates an interpolation value from the adjacent original image data (step S1). The filter 4 receives the original image data and the interpolation data output from the upsampling unit 3 and performs a predetermined filter process (S2).
[0019]
The original image data and the interpolation data after the upsampling process and the filter process are input to the minority processing unit 5 to perform decimal point processing of the interpolation data (step S3). Here, the interpolation data f (x) generated for the adjacent pixel values x1 and x2 can be expressed by (x1 + x2) / 2 as shown in FIG. In the actual decimal point processing, the mathematical expression is obtained by filtering the mathematical expression. However, for the sake of simplicity, the decimal point processing will be described using interpolation data of the mathematical expression.
[0020]
When the interpolated data f (x) includes a fractional part, the decimal processing unit 5 performs a truncation process after the decimal point and outputs the information data from the information data unit 7 to the data multiplexing unit. To do. Further, when the decimal point is not included, the information is output to the subsequent stage as it is and the information data portion 7 is not multiplexed. Here, the case where the fractional value of the interpolation data f (x) does not become 0 is limited to the case where the combination of x1 and x2 is an odd number and an even number. If x1 is an even number and x2 is an odd number, and x1 = 2p and x2 = 2q + 1 (p, q non-negative integer), then f (x) = (x1 + x2) / 2. p + q + 0.5.
[0021]
The data multiplexing unit 6 adds and outputs the information data read from the information data unit 7 to the LSB (integer part) of the interpolation data only when the decimal part is rounded down as a result of the processing in the decimal processing unit 5. (Step S4). However, when the number of bits of the original image data is n, only when the interpolated data exceeds 2 ⁿ −1, even if there is a decimal point, the information data is not multiplexed, and the decimal part is truncated. In the above case, when the pixel data is 255.0 or more, the information data multiplexing process is not performed.
[0022]
As described above, the error between the pixel value and the true value obtained by data multiplexing is less than 1 because 0 or 1 is added after the decimal part is rounded down.
[0023]
Next, the operation of extracting data received from the data multiplexing device on the receiving side will be described. FIG. 2 shows a block diagram of the data extraction apparatus according to the present embodiment. The data extraction device 2 includes an interpolation data extraction unit 12 that extracts interpolation data from the received multiple image data, a down sample unit 8 that extracts original image data from the multiple image data by down sampling, and the extracted original image data. An upsampling unit 3 that reproduces interpolation data by upsampling, a filter unit 4 that filters the output of the upsampling unit 3, and a small number that determines whether or not a decimal point is included by referring to the reproduced interpolation data after filtering. Presence / absence determination unit 9, data integer part LSB comparison unit 10 that compares the interpolation data extracted from the multiplexed image data and the reproduction interpolation data in response to this determination, and information data extraction that extracts information data according to the comparison result Unit 11.
[0024]
Next, the operation of extracting information data multiplexed on multiple image data will be described using the flowchart shown in FIG. In the following description, an operation for extracting one bit of information data will be described.
[0025]
When receiving the multiple image data, the down-sampling unit 8 extracts the original image data by down-sampling (step S11).
[0026]
The up-sampling unit 3 reproduces the interpolation data from the extracted adjacent original image data in the same manner as the transmission side, and outputs this as reproduction interpolation data (step S12).
[0027]
The minority presence / absence determining unit 9 determines whether or not the reproduction interpolation data subjected to the filtering process includes a fractional part. Here, it is assumed that “0” is output as the determination result when the reproduction interpolation data does not include a value after the decimal point, and “1” is output when the reproduction interpolation data includes a value after the decimal point. When the determination result “1” is output, the reproduction interpolation data is also output to the subsequent stage.
[0028]
When the reproduced interpolation data includes a decimal point, it can be seen that, as described above, the transmission side has been subjected to the truncation process below the decimal point and the information data multiplexing process. Therefore, when the determination result “1” is output from the minority presence / absence determination unit 9, the data integer part LSB comparison unit 10 extracts information data from the interpolation data from the interpolation data extraction unit 12 according to the following procedure.
[0029]
The data integer part LSB comparison unit 10 compares the LSB of the interpolation data output from the interpolation data extraction unit 12 with the LSB of the integer part in the reproduction interpolation data received from the decimal number determination unit 9. As a result, if both match, it is determined that information data “0” is multiplexed on the transmission side, and “0” is output. If they do not match, the transmission side determines that the information data “1” is multiplexed and outputs “1”.
[0030]
The comparison between the two can be obtained by exclusive OR (EXOR) (step S15). That is, if the result of the EXOR operation is 0 (LSB matches), “0” is output as information data (step S16). On the other hand, in the case of 1 (LSB does not match), “1” is output as information data (step S17).
[0031]
In the above description, the method of multiplexing and extracting 1-bit information data has been described. In general, header information including the number of multiplexed data is added to the information data. Based on this, the procedure for continuously multiplexing and extracting information data is clear.
[0032]
The present invention can be applied not only to the field of broadcasting but also to image enlargement on a network such as the Internet.
[0033]
【The invention's effect】
As described above, according to the present invention, it is possible to reduce the deterioration of the image that occurs when the original image is enlarged and the data multiplex processing is performed on the enlarged image. This is because, in the present invention, when the interpolation data has a decimal point, the information is cut off and the error is reduced as compared with the conventional case. Also, when the decimal point is not included in the interpolation data, the information data is not inserted, so that the original image data is not deteriorated.
[Brief description of the drawings]
FIG. 1 is a block diagram of a data multiplexing apparatus according to an embodiment of the present invention.
FIG. 2 is a block diagram of a data extraction apparatus according to an embodiment of the present invention.
FIG. 3 is a block diagram of a conventional data multiplexing apparatus.
4 is a flowchart showing a data multiplexing method of the data multiplexing apparatus 1. FIG.
FIG. 5 is a flowchart showing a data extraction method of the data extraction device 2;
FIG. 6 is a conceptual diagram of linear interpolation when the original image is doubled.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 Data multiplexing apparatus 2 Data extraction apparatus 3 Upsampling part 4 Filter part 5 Decimal part 6 Data multiplexing part 7 Information data 8 Downsampling part 9 Decimal presence judgment part 10 Data integer part LSB comparison part 11 Data extraction part 12 LSB processing part 13 Interpolation data extraction unit

Claims (7)

Interpolation data for the adjacent image data is obtained from the adjacent original image data based on the original image data, and when the interpolation data includes a value after the decimal point, decimal point processing is performed to round off the value after the decimal point. Information data is added to the least significant bit of the interpolation data only for the interpolation data that has been processed, and each of the interpolation data with or without decimal point processing inserted between the adjacent original image data A method for multiplexing data of information data with respect to enlarged image data. Multiplex image data multiplexed by the data multiplexing method is received, original image data is generated by downsampling, reproduction interpolation data is obtained from adjacent original image data based on the original image data, and the received multiple image data When the reproduced interpolation data includes a value after the decimal point, the interpolation data is compared with the least significant bit in the integer part of the reproduced interpolation data. A data extraction method for information data with respect to an enlarged image, characterized in that a logical value of 1 is output and, if they do not match, a 1-level logical value is output as information data. A non-transitory computer-readable storage medium storing a program for causing a computer to execute a data multiplexing method for information data with respect to enlarged image data according to claim 1. A non-transitory computer-readable storage medium storing a program for causing a computer to execute the data extraction method for information data with respect to the enlarged image data according to claim 2. An upsampling unit for obtaining interpolation data for the adjacent image data from adjacent original image data based on the original image data;
When the interpolation data includes a value after the decimal point, a decimal processing unit that rounds off the value after the decimal point,
Information data is added to the least significant bit of the interpolation data only for the interpolation data obtained by rounding down the value after the decimal point, and each of the interpolated data subjected to the decimal point processing or the interpolation data not subjected to the decimal point processing is adjacent to the original image data. A data multiplexing part to be inserted between
A data multiplexing apparatus for information data with respect to enlarged image data. A downsampling unit that receives multiplexed image data multiplexed by the data multiplexing device and generates original image data by downsampling;
An upsampling unit that generates reproduction interpolation data from adjacent original image data based on the original image data;
An interpolation data extraction unit for extracting interpolation data from the received multiple image data;
A minority presence / absence determining unit that determines whether or not the reproduction interpolation data includes a value after the decimal point;
When it is determined by the determination that the reproduction interpolation data includes a value after the decimal point, the least significant bit in the integer part of the reproduction interpolation data and the reproduction interpolation data are compared. A data integer part LSB comparison unit that outputs a logical value and outputs a one-level logical value as information data in the case of mismatch,
A data extraction device for information data with respect to enlarged image data. 7. The data extraction device for information data with respect to enlarged image data according to claim 6, wherein the comparison in said data integer part LSB comparison part is performed by an exclusive OR circuit.

Images