JP3996176B2 - Image data processing apparatus, image data processing method, and image data processing program - Google Patents

Description

  The present invention relates to an image data processing apparatus, an image data processing method, and an image data processing program for extracting a code embedded in image data, and in particular, reduces processing required for decoding a code from image data. The present invention relates to an image data processing apparatus, an image data processing method, and an image data processing program.

  As disclosed in U.S. Pat. No. 5,636,292 (Patent Document 1) and Japanese Patent Application Laid-Open No. 2000-299779 (Patent Document 2), data other than image data and audio data has been conventionally used. The technique of embedding (code, etc.) is applied to prevent counterfeiting, prevent unauthorized use, and provide additional services.

  As described above, since the application of such a technique is security, conventionally, only a method capable of withstanding the transformation of the original data and the request for partial use is taken. For example, in the past, a very complicated method has been taken (such as digital watermark technology), in which the same data is dispersedly arranged in an image or data is input in the frequency domain using FFT (Fast Fourier Transform).

  Here, the digital watermark technology is applied to various additional services. For example, U.S. Pat. No. 5,841,978 (Patent Document 3) discloses a method of reading a digital watermark embedded in a printed material and displaying a specific Web page.

U.S. Pat. No. 5,636,292 JP 2000-299779 A U.S. Pat. No. 5,841,978

  Incidentally, as described above, in the conventional digital watermark technique, it is necessary to perform FFT calculation in order to enter data in the frequency domain, but the calculation amount required for FFT is enormous.

  For this reason, in portable information devices (such as mobile phones, PHS (Personal Handyphone System), PDA (Personal Digital Assistant), etc.) whose processing performance of computer resources such as memory and processor is limited, data is embedded and read out from image data. There is a problem that it is difficult to carry out a practical processing time.

  The present invention has been made in view of the above, and an object of the present invention is to provide an image data processing device, an image data processing method, and an image data processing program capable of reducing the processing required for decoding a code from image data. To do.

  In order to achieve the above object, the present invention provides an image data processing apparatus for extracting a code embedded in image data, a dividing means for dividing the image data into a plurality of blocks, and each block of the plurality of blocks. And a decoding means for extracting a code from the pair block based on the magnitude relationship of the feature quantity between each block of a pair block consisting of two blocks. When the feature amount difference between the blocks in the pair block is not equal to or greater than an upper threshold value, and the feature amount of the right block in the pair block is larger than the feature amount of the left block, the bit from the pair block The first code is extracted as a determination result, and the difference in feature amount between the blocks in the pair block is When the feature value of the left block in the pair block is greater than the limit threshold and the feature value of the right block is larger than the feature value of the right block, a second code is extracted from the pair block as a bit determination result. If the difference in feature quantity between the blocks is equal to or greater than an upper threshold, a third code is extracted from the pair block as being uncertain, and the first code, the second code, and A code is extracted based on the third code.

  According to another aspect of the present invention, there is provided an image data processing method for extracting a code embedded in image data, a dividing step of dividing the image data into a plurality of blocks, and a feature amount in each block of the plurality of blocks. A feature amount extracting step, and a decoding step for extracting a code from the pair block based on the magnitude relationship of the feature amounts between the blocks of a pair block consisting of two blocks, wherein the decoding step includes the pair block If the feature amount difference between the blocks in the pair block is not equal to or greater than the upper threshold value and the feature amount of the right block in the pair block is larger than the feature amount of the left block, the first bit determination result from the pair block is The code is extracted, and the difference in feature amount between the blocks in the pair block is not more than the upper threshold value. When the feature quantity of the left block in the pair block is larger than the feature quantity of the right block, a second code is extracted from the pair block as a bit determination result, and the difference in the feature quantity between the blocks in the pair block is the upper limit. If it is greater than or equal to a threshold value, a third code is extracted from the pair block as an uncertain bit determination result, and a code based on the first code, the second code, and the third code is extracted. Is extracted.

  According to another aspect of the present invention, there is provided an image data processing program for extracting a code embedded in image data. The computer includes a dividing unit that divides the image data into a plurality of blocks, and a feature amount in each block of the plurality of blocks. Extracting feature quantity extracting means, functioning as decoding means for extracting code from the pair block based on the magnitude relation of the feature quantity between the blocks of a pair block consisting of two blocks, the decoding means comprising the pair block If the feature amount difference between the blocks in the pair block is not equal to or greater than the upper threshold value and the feature amount of the right block in the pair block is larger than the feature amount of the left block, the first bit determination result from the pair block is The code is extracted, and the difference in feature quantity between the blocks in the pair block is the upper limit. When the feature amount of the left block in the pair block is not greater than the threshold value and greater than the feature amount of the right block, a second code is extracted as a bit determination result from the pair block, and the block between the pair blocks If the difference between the feature quantities is equal to or greater than an upper threshold, a third code is extracted from the pair block as being uncertain, and the first code, the second code, and the first code are extracted. A code is extracted based on the three codes.

  According to this invention, the image data is divided into a plurality of blocks, the feature amount in each block of the plurality of blocks is extracted, and the difference in the feature amount between the blocks in the pair block consisting of two blocks is equal to or greater than the upper threshold value. If the feature amount of the right block in the pair block is larger than the feature amount of the left block, the first code is extracted as the bit determination result from the pair block, and the difference in feature amount between the blocks in the pair block is When the feature value of the left block in the pair block is greater than the upper threshold value and the feature value of the left block is larger than the feature value of the right block, the second code is extracted from the pair block as a bit determination result, and between the blocks in the pair block If the feature value difference between the pair is greater than or equal to the upper threshold, the bit decision result from the pair block is indeterminate. As the third code is extracted and the code is extracted based on the first code, the second code, and the third code, it is possible to prevent image quality deterioration due to excessive feature amount change. it can.

  According to the present invention, when the difference in feature quantity between blocks in a pair block exceeds the upper threshold, the code is not extracted from the pair block, so image quality deterioration due to excessive feature quantity change. There is an effect that can be prevented.

  Further, according to the present invention, since the error correction of the code embedded in the image is performed, the error correction can be performed and the reliability can be improved.

  Further, according to the present invention, since the feature amount in a part of each block is extracted, the processing amount can be reduced as compared with the case where the entire feature amount of the block is used.

  In addition, according to the present invention, since the feature amount in the central portion of each block is extracted, the processing amount can be reduced as compared with the case where the entire feature amount of the block is used. .

  Further, according to the present invention, the feature amount of the specific color component included in the image data is extracted, and the code embedded in the plurality of blocks is extracted based on the feature amount of the extracted specific color component. Even if the characteristic amount of the specific color component is changed, it is possible to maintain the data discrimination ability without degrading the image quality by using the characteristic that it is difficult to distinguish with the human eye.

  In addition, according to the present invention, the feature amount of the yellow component included in the image data is extracted, and the code embedded in a plurality of blocks is extracted based on the extracted feature amount of the yellow component. By utilizing the inconspicuous property, the data discrimination ability can be maintained without degrading the image quality.

  In addition, according to the present invention, since the image is divided based on the cutout cutout mark pixels added to the image, there is an effect that the image can be cut out easily.

  In addition, according to the present invention, the feature amount is at least one of average density, granularity, saturation, density centroid or dispersion, so that the feature quantity is not noticeable and the influence on the image quality can be reduced. There is an effect that can be done.

  Embodiments 1 to 5 of an image data processing apparatus, an image data processing method, and an image data processing program according to the present invention will be described below in detail with reference to the drawings.

(Embodiment 1)
In the following, as a specific example of the image data processing apparatus according to the first embodiment, an encoder 100 (see FIG. 1) for generating a coded image data by embedding a code (binary number) as another data in original image data. And a decoder 200 (see FIG. 9) that reads a code from the coded image data will be described.

  FIG. 1 is a block diagram showing a configuration of an encoder 100 according to the first embodiment of the present invention. The encoder 100 is a device for embedding a code C (see FIG. 4) eight times, for example, in the original image data I0 (see FIG. 2) to generate image coded data I3 (see FIG. 5).

  The original image data I0 is image data generated in a predetermined format (JPEG (Joint Photographic Expert Group), GIF (Graphics Interchange Format, etc.)), and is 1024 × 1024 pixel size image data. For example, a 16-bit code C (1010110101001010) shown in FIG. 4 is embedded in the original image data I0.

  In the encoder 100, the block dividing unit 101 divides the input original image data I0 into blocks of N rows × M columns (16 × 16 in the case of FIG. 3) as shown in FIG. Output as divided image data I1.

The block-divided image data I1 is the block _{B l11, B r11, ···,} B l18, B r18, B l21, B r21, ···, consist B _L168, B _R168 that 256 (16 × 16) blocks Has been. One block has a size of 64 × 64 pixels.

  Here, in the block-divided image data I1, a 1-bit code is embedded in a pair block (two adjacent blocks).

Specifically, the pair of blocks, the block B _l11 and B _r11, block B _l12 and B _r12, · · ·, blocks B _l18 and B _r18 (1 row far), the block B _l21 and B _r21, · · _.. , Blocks B _l28 and B _r28 (the second row so far),..., Blocks B _l161 and B _r161 ,..., Blocks B _l168 and B _r168 (the 16th row so far) Has been.

Here, in one block B _lxy of the pair block, the subscript l indicates that it is the left block in the pair block. The subscript x represents a row (N). The subscript y represents the column (M). On the other hand, in the block B _rxy of the pair block, the subscript r represents the right block in the pair block. The subscript x represents a row (N). The subscript y represents the column (M).

In the pair block, the average density (average gradation of each pixel in the block) as the feature quantity in the left block B _lxy is set as the left average density data D _l, and the average density (feature quantity) in the right block B _rxy is used. Is the right average density data _Dr.

Here, as shown in the following relational expression, when the left average density data D _l is less than the right average density data D _r , the pair block represents “0” as a one-bit code. On the other hand, when the left average density data D _l is equal to or greater than the right average density data D _r , the pair block represents “1” as a one-bit code.

D _l <D _r → “0”
D _l ≧ D _r → “1”

For example, the pair block consisting of blocks B _l18 and B _r18 as shown in FIG. 3, the left average density data D _l18 is "115", since the right average density data D _r18 is "125", one-bit code of Represents “0”.

In the pair block consisting of blocks B _l28 and B _r28 , the left average density data D _l28 is “125” and the right average density data D _r28 is “115”, so that “1” is represented as a one-bit code. .

  Further, since the block-divided image data I1 has 8 pair blocks (16 blocks) per row, it represents an 8-bit code. Therefore, all lines (16 lines) represent 128-bit codes. In Embodiment 1, since the code C (see FIG. 4) embedded in the block divided image data I1 is 16 bits, the code C can be embedded in the block divided image data I1 up to 8 (128/16) times (see FIG. 4). (See FIG. 5).

Returning to FIG. 1, the block extraction unit 102 sequentially extracts the paired blocks (block B _lxy and block B _rxy ) from the block divided image data I1 (see FIG. 3) by following the bit shift of the code C, and blocks B _lxy The density distribution in each of the blocks _Brxy is sequentially output as block density data D.

  Here, the bit shift of the code C means that a bit pointer (not shown) is shifted to the right by one bit from the leftmost bit (1) shown in FIG. 4 to the right bit (0). .

The averaging unit 103 _obtains left average density data D _l corresponding to the block B _lxy and right average density data D _r corresponding to the block B _{rxy from} the block density data D, and these are obtained in the register 104 _l and the register 104. _{The r} is sequentially stored following the bit shift of the code C.

Comparing section 105, n bits of the code C (n = 1, 2 from the leftmost bit shown in FIG. 4, ..., 16) and the left average density stored in the registers 104 _l and the register 104 _r A bit determination result determined from the magnitude relationship between the data D _l and the right-side average density data D _r (bit determination of “0” or “1” by the relational expression described above: see FIG. 3) is made.

Based on the comparison result of the comparison unit 105, the encoding unit 106 executes a process for embedding the code C in the block divided image data I1 (original image data I0). Specifically, when the comparison result of the comparison unit 105 is the same, the encoding unit 106 maintains the magnitude relationship between the left average density data D _l and the right average density data _Dr , while the comparison result is inconsistent. In some cases, the left side average density data D _l and the right side average density data _Dr are changed (the magnitude relation is reversed) so as to have a magnitude relationship representing the bits of the code C, and the image encoded data I3 (see FIG. 5) is generated. And output this.

Image coded data I3 shown in FIG. 5, block-divided image data I1 corresponds to (see FIG. 3) and the original image data I0 (refer to FIG. 2), has an area A ₁ to A _8. In the areas A _{1 to} A ₈ , the same code C (1010110101001010) is embedded eight times in total.

For example, the area A ₁ is block B _l11 shown in FIG. 3, B _r11, · · ·, and corresponds to the B _l28, B _r28. The other areas A _{2 to} A ₈ correspond to the blocks B _l31 , B _r31 ,..., B _l168 , B _r168 .

  In FIG. 5, the embedded state of the code C (see FIG. 3) is shown, but the actual image coded data I3 is almost the same as the original image data I0 (see FIG. 2) (partial density is changed). In some cases, the image data may be invisible to the naked eye.

  Each component of the encoder 100 is interconnected via a control unit (not shown).

  Next, the operation of the encoder 100 shown in FIG. 1 will be described with reference to the flowcharts shown in FIGS. FIG. 6 is a flowchart for explaining an operation example 1 of the encoder 100.

  In the figure, in step SA1, a code C (see FIG. 3) is set in the comparison unit 105. In step SA2, the comparison unit 105 sets n to 1 as initialization. This n represents a bit pointer of code C as described above. In this case, n = 1 corresponds to the leftmost bit (“1”) of the code C.

In step SA3, the original image data I0 (see FIG. 2) is input to the block dividing unit 101. At step SA4, the block division unit 101, the block division processing, the input original image data I0 is divided into blocks B _l11 .about.B _R168 of the 16 × 16 as shown in FIG. 3, which block division image data It is output to the block extraction unit 102 as I1.

In step SA5, the block extraction unit 102 extracts a pair block (in this case, a block _B11 and a block _Br11 ) corresponding to n = 1 from the block divided image data I1, and then the block _B11 and the block _Br11 . Each density distribution is output to the averaging unit 103 as block density data D.

At step SA6, the averaging unit 103, the averaging process, from the block density data D, the left average density data D _l11 corresponding to the block B _l11 and (not shown), the right-side average density data D corresponding to the block B _{r11 r11} (not shown) is obtained.

In step SA7, the averaging unit 103, the left-side average density data D _l11 (not shown) in the register 104 _l, respectively stored right-side average density data D _r11 (not shown) in the register 104 _r.

At step SA8, comparing unit 105 is a leftmost bit (corresponding to n = 1) of the code C shown in FIG. 4 as "1", the left average density data stored in the registers 104 _l and the register 104 _r A density difference between _D11 and right average density data _Dr11 is obtained, and bit determination is performed from the density difference (magnitude relationship).

In this case, _assuming that the left average density data _D11 is greater than or equal to the right average density data _Dr11 , the comparison unit 105 sets the bit determination result in the pair block to “1” based on the above-described magnitude relationship.

  In step SA9, the comparison unit 105 determines that the nth bit of the code C (in this case, the first bit is “1”) is the same as the bit determination result in step SA8 (in this case, “1”). In this case, the determination result is “Yes”.

  In step SA10, the comparison unit 105 increments n by 1. As a result, n is set to 2. In step SA11, the comparison unit 105 determines whether n is greater than nend. nend is the total number of bits of the code C (see FIG. 4) and is 16. In this case, since n is 2, the comparison unit 105 sets “No” as the determination result of step SA11.

At step SA5, the block extraction unit 102, the block division image data I1, a pair of blocks (in this case, the block B _l12 and the block B _r12) corresponding to n = 2 after extracting, the blocks B _l12 and the block B _r12 Each density distribution is output to the averaging unit 103 as block density data D.

At step SA6, the averaging unit 103, the averaging process, from the block density data D, the left average density data D _l12 corresponding to the block B _l12 and (not shown), the right-side average density data D corresponding to the block B _{r12 r12} (not shown) is obtained.

In step SA7, the averaging unit 103, the left-side average density data D _l12 (not shown) in the register 104 _l, respectively stored right-side average density data D _r12 (not shown) in the register 104 _r.

At step SA8, comparing section 105, a next bit in the code C shown in FIG. 4 (corresponding to n = 2) and "0", the left average density data stored in the registers 104 _l and the register 104 _r D _l12 and determine the concentration difference of the right average density data D _r12, performs bit decision from the density difference (magnitude relationship).

In this case, when the left average density data D _l12 is to be less than the right average density data D _r11, comparing unit 105, a bit determination result in the pair of blocks from the magnitude relation as described above to "0".

  In step SA9, the comparison unit 105 determines that the nth bit of code C (in this case, the second bit is “0”) and the bit determination result in step SA8 (in this case “0”) are the same. In this case, the determination result is “Yes”.

  In step SA10, the comparison unit 105 increments n by 1. As a result, n is set to 3. In step SA11, the comparison unit 105 determines whether n (= 3) is greater than nend (= 16). In this case, the determination result is “No”. Thereafter, the above-described operations after step SA5 are repeated until the determination result at step SA11 becomes “Yes”.

When n is set to 16 in step SA10 and the determination result in step SA11 is “No”, in step SA5, the block extraction unit 102 determines the pair block corresponding to n = 16 from the block divided image data I1. (In this case, after extracting the block B _l28 and the block B _r28 ), the density distribution in each of the block B _l28 and the block B _r28 is output to the averaging unit 103 as the block density data D.

At step SA6, the averaging unit 103, the averaging process, from the block density data D, the left average density data D _l28 corresponding to the block B _l28: (the "125" refer to FIG. 3), corresponding to the block B _r28 Right average density data D _r28 (“115”: see FIG. 3) is obtained.

In step SA7, the averaging unit 103 stores the left average density data D _l28 (“125”) in the register 104 _l and the right average density data D _r28 (“115”) in the register 104 _r .

At step SA8, comparing section 105, a (corresponding to n = 16) right bit of the code C shown in FIG. 4 as "0", the register 104 _l and the register 104 left average density stored in the _r data D A density difference (10) between _l28 (“125”) and right-side average density data D _r28 (“115”) is obtained, and bit determination is performed from the density difference (magnitude relationship).

In this case, since the left average density data D _l28 (“125”) is equal to or greater than the right average density data D _r28 (“115”), the comparison unit 105 determines the bit determination result in the pair block as “ 1 ”.

  In step SA9, the comparison unit 105 determines that the nth bit of the code C (in this case, the 16th bit is “0”) and the bit determination result in step SA8 (in this case, “1”) are the same. In this case, the determination result is “No”.

  In step SA14, the comparison unit 105 determines whether or not the density difference (10) obtained in step SA8 is equal to or less than a preset upper threshold (for example, 100). In this case, the determination result Is “Yes”.

In step SA15, the encoding unit 106 determines that the bit determination result based on the magnitude relationship between the left average density data D _l28 (“125”) and the right average density data D _r28 (“115”) is the nth bit of the code C ( In this case, density change processing is executed in which the left average density data D _l28 and the right average density data D _r28 are changed so as to be the same as “0” in the 16th bit.

That is, the encoding unit 106 reverses the magnitude relationship between the left average density data D _l28 (“125”) and the right average density data D _r28 (“115”), thereby _{converting the} left average density data D _l28 into the right average density data. The density is changed so that the bit determination result changes from “1” to “0”, _{assuming that the} data is less than the data _{Dr 28} .

Specifically, the encoding unit 106 changes the left-side average density data D ′ _l (changed left-side average density data from the expression (1) when (A) D _l <D _r shown in FIG. 'after obtaining the corresponding) to _l28, (2) the right average density data D after the change from the equation' D seek corresponding to the right-side average density data D _'r28 after _r (changed).

Thereby, after the density change, the left average density data D ′ _l28 becomes less than the right average density data D ′ _r28 , and the bit determination result is changed from “1” to “0”.

On the other hand, when (B) D _l ≧ D _r shown in FIG. 7, the left average density data D ′ _l after the change from the equation (3) (corresponding to the left average density data D ′ _lxy after the change) Is obtained, the changed right average density data D ′ _r (corresponding to the changed right average density data D ′ _rxy ) is obtained from the equation (4).

Thereby, after the density change, the left average density data D ′ _lxy becomes equal to or higher than the right average density data D ′ _rxy , and the bit determination result is changed from “0” to “1”.

  If the determination result in step SA14 is “No”, that is, if the density difference obtained in step SA8 is greater than a preset upper threshold (for example, 100), the density change process The process of step SA10 is executed without executing.

  This is because if the density difference in the pair block is large, if density change processing is executed, it will be known that the appearance has been changed and the image quality will be deteriorated (unnatural image). Is not done.

  In step SA10, the comparison unit 105 increments n by 1. As a result, n is set to 17. In step SA11, the comparison unit 105 determines whether n (= 17) is greater than nend (= 16). In this case, the determination result is “Yes”.

In step SA12, the comparison unit 105 determines whether or not the above-described processing relating to the final pair block (blocks _Bl168 and _Br168 ) in the block-divided image data I1 has been completed. In this case, the determination result is “No”. And

In step SA16, the comparison unit 105 sets n to 1 to reset n (= 17). In step SA5, the block extraction unit 102 extracts the next pair block (in this case, block B _l31 and block B _r31 : see FIG. 3) corresponding to n = 1 from the block divided image data I1, and then blocks B _l31 and density distributions of the block B _r31 outputs to the averaging section 103 as the block density data D.

  Thereafter, the above-described operation is repeated until the determination result in step SA12 is “Yes”.

  When the determination result in step SA12 is “Yes”, in step SA13, the encoding unit 106 determines that the encoded image data I3 is based on the determination result in step SA9, the determination result in step SA14, and the density change process in step SA15. Is generated and output.

Specifically, the encoding unit 106 determines the left average density data D _l and the right average density data D _r for the pair block for which the determination result in step SA9 is “Yes” (the determination result in step SA14 is “No”). On the other hand, based on the density change process in step SA15, the coded image data corresponding to the changed left average density data D' _l and right average density data D' _r is maintained. I3 is generated. This encoded image data I3 is decoded by a decoder 200 described later.

Here, the same code C (1010110101001010) is embedded eight times in total in the areas A _{1 to} A ₈ in the image encoded data I3 shown in FIG.

  Next, an operation example 2 of the encoder 100 shown in FIG. 1 will be described with reference to the flowchart shown in FIG. In FIG. 8, step SB10 and step SB11 are newly added.

  Therefore, step SB1 to step SB9 and step SB12 to step SB18 shown in FIG. 8 correspond to step SA1 to step SA16 shown in FIG.

  In step SB9 shown in FIG. 8, the comparison unit 105 determines whether or not the nth bit of the code C is the same as the bit determination result in step SB8. In this case, the determination result is “Yes”. And

  In step SB10, the comparison unit 105 determines whether or not the density difference obtained in step SB8 is less than a preset lower threshold (for example, 10). In this case, the determination result is “Yes”.

  Here, when the density difference is less than the lower limit threshold, the accuracy is lowered, for example, the magnitude relationship is reversed during decoding.

In order to avoid such a problem, in step SB11, encoding section 106, as the density difference is greater than or equal to the lower threshold, and adjusting the left-side average density data D _l and the right average density data D _r, the density difference The density difference enlargement process for enlarging the image is executed.

  If the determination result in step SB10 is “No”, that is, if the density difference is greater than or equal to the lower threshold value, the process in step SB12 is executed. Also, when the determination result of step SB16 is “No”, the process of step SB12 is executed.

  FIG. 9 is a block diagram showing a configuration of the decoder 200 according to the first embodiment of the present invention. The decoder 200 is a device for decoding an embedded code from the image coded data I3 (see FIG. 5) encoded by the encoder 100 (see FIG. 1).

  In the decoder 200, when the image cutout unit 201 includes image data (for example, a blank portion) around the image coded data I3, the image cutout unit 201 has a function of cutting out valid image coded data I3 from the whole. However, when only the image coded data I3 is input to the image cutout unit 201, the cutout is not performed.

  In the same way as the block division image data I1 shown in FIG. 3, the block division unit 202 converts the image coded data I3 from the image cutout unit 201 into N rows × M columns (in the case of FIG. 3, 16 × 16). And are output as block divided image data (not shown).

  In the same manner as the block extraction unit 102 (see FIG. 1), the block extraction unit 203 sequentially causes the pair blocks (two blocks) from the block divided image data to follow the bit shift of the decoded code (16 bits). Extracted and the density distribution in the pair block (two blocks) is sequentially output as block density data (not shown).

From the block density data, the averaging unit 204 corresponds to the left average density data (not shown) corresponding to one block in the pair block and the other block in the same manner as the averaging unit 103 (see FIG. 1). Right average density data (not shown) is obtained, and these are sequentially stored in the register 205 _l and the register 205 _r following the bit shift of the code.

The comparison unit 206 performs bit determination by comparing the magnitude relationship between the left average density data and the right average density data stored in the register 205 _l and the register 205 _r , and determines the bit determination result (“ A code group CG (candidate codes C _{1 to} C ₈ : see FIG. 10) corresponding to “0” or “1” is output to the decoding unit 207.

Here, each of the candidate codes C _{1 to} C ₈ shown in FIG. 10 has a 16-bit configuration, and each code embedded in the areas A _{1 to} A ₈ of the image coded data I3 (see FIG. 5). (16 bits) is a result of decoding, and is a candidate for a code C ′ (see FIG. 9) as a decoding result of the decoder 200.

In the candidate codes C _{1 to} C ₈ , “2” represents a bit whose bit determination of “1” or “0” is indeterminate.

As shown in FIG. 10, the decoding unit 207 takes a majority vote in bit units (bit string in the vertical direction in the figure) from the candidate codes C _{1 to} C ₈ corresponding to the comparison result of the comparison unit 206, and each bit (all 16 Bit) is determined, and this is output as a code C ′ as a decoding result of the decoder 200.

  Each component of the decoder 200 is interconnected via a control unit (not shown).

  Next, the operation of the decoder 200 shown in FIG. 9 will be described with reference to the flowchart shown in FIG.

  In the figure, in step SC1, image coded data I3 (see FIG. 1) is input to the image cutout unit 201. In step SC2, n is set to 1 as initialization. This n represents a bit pointer of the code to be decoded. In this case, n = 1 corresponds to the leftmost bit of the code.

  In step SC3, when the image cutout unit 201 includes image data (for example, a blank portion) around the input image coded data I3, the image cutout unit 201 cuts out valid image coded data I3 from the whole.

  In step SC4, the block dividing unit 202 converts the image coded data I3 extracted by the image extracting unit 201 into 16 × 16 blocks by the block dividing process in the same manner as the block dividing unit 101 (see FIG. 1). This is divided and output to the block extraction unit 203 as block divided image data (not shown).

  In step SC5, the block extraction unit 203 extracts a pair block (two blocks) corresponding to n = 1 from the block divided image data (not shown) in the same manner as the block extraction unit 102 (see FIG. 1). Thereafter, the density distribution in each block is output to the averaging unit 204 as block density data.

  In step SC6, the averaging unit 204, as with the averaging unit 103 (see FIG. 1), performs averaging processing from the block density data to the left average density data (not shown) corresponding to one block, The right average density data (not shown) corresponding to the other block is obtained.

In step SC7, the averaging unit 204, the left-side average density data (not shown) in the register 205 _l, respectively store the right average density data (not shown) in the register 205 _r.

In step SC8, comparing unit 206, by comparing the left average density data and the magnitude relation between the right average density data stored in the registers 205 _l and the register 205 _r, performs bit decision, the bit determination result (the above-mentioned The bit is determined as “0” or “1” according to the relational expression), and is output to the decoding unit 207.

  Here, the comparison unit 206 obtains a density difference between the left average density data and the right average density data, and if this density difference is larger than a certain upper threshold value, a bit determination result corresponding to the pair block. Is unreliable, and the bit determination result is “2” (uncertain: see FIG. 10).

  In step SC9, the comparison unit 206 increments n by 1. As a result, n is set to 2. In step SC10, the comparison unit 206 determines whether n is greater than nend (= 16). In this case, the determination result is “No”.

  Thereafter, the operation after step SC5 is repeated until the determination result of step SC10 is “Yes”.

When n is set to 17 in step SC9, the determination result in step SC10 is “Yes”. At this point, the comparison unit 206 uses the candidate code C ₁ shown in FIG. 10 as the bit discrimination result.

  In step SC11, the comparison unit 206 determines whether or not the above-described processing relating to the last pair block (two blocks) in the block-divided image data (not shown) has been completed. In this case, the determination result is “No”. And

  In step SC14, the comparison unit 206 sets n to 1 to reset n (= 17). In step SC5, the block extraction unit 203 extracts the next pair block (two blocks) corresponding to n = 1 from the block divided image data (not shown), and then uses the density distribution in each block as block density data. The result is output to the averaging unit 204.

  Thereafter, the above-described operation is repeated until the determination result in step SC11 becomes “Yes”.

When the determination result in step SC11 is “Yes”, in step SC12, the decoding unit 207 executes majority decision processing. That is, at this time, the candidate codes C _{1 to} C ₈ shown in FIG. 10 are the bit determination results.

The decoding unit 207 takes a majority vote in bit units (bit string in the vertical direction in the figure) from the candidate codes C _{1 to} C ₈ and determines each bit (all 16 bits). For example, in the case of the leftmost bits of the candidate codes C _{1 to} C ₈ , since “0” is 2, “1” is 5, and “2” is 1, the leftmost bit of the code C ′ is determined by the majority decision. “1” is confirmed.

  In step SC13, the decoding unit 207 receives the result of the majority decision process and outputs a code C ′ (see FIG. 10) as a decoding result of the decoder 200. The code C ′ is the same as the code C (see FIG. 4).

  As described above, according to the first embodiment, an average density (feature) is obtained for each pair block in a plurality of blocks (block divided image data I1: see FIG. 3) obtained by dividing the original image data I0 (see FIG. 2). Since a single bit (code) is associated with each other based on the magnitude relation of the amount) and the code C (plural codes) is embedded in a plurality of blocks, the conventional FFT is not required and the code is embedded in the image data. The processing required for this can be reduced.

  Further, according to the first embodiment, as described in the operation example 1 (see FIG. 6), when the average density magnitude relationship in the pair block does not match the bit (code) to be embedded, the magnitude relationship is reversed. As described above, since the average density in the pair block is changed, an arbitrary code can be freely embedded.

  Further, according to the first embodiment, as described in the operation example 1 (see FIG. 6), when the density difference between the blocks in the pair blocks (difference in feature amount) exceeds the upper limit threshold value, Since the average density is not changed, it is possible to prevent image quality deterioration due to an excessive change in the average density.

  Further, according to the first embodiment, as described in the operation example 2 (see FIG. 8), the magnitude relationship of the average density (feature amount) matches the bit (code) to be embedded, and the block in the pair block When the difference in density (difference in feature amount) is less than the lower threshold value, the average density is changed so that the difference becomes equal to or higher than the lower threshold value. It is possible to prevent a decrease in accuracy, such as a reversal of the magnitude relationship.

  Further, according to the first embodiment, since the feature amount is set to the average density, the calculation related to the feature amount can be simplified.

  Also, according to the first embodiment, as shown in FIG. 5, since the code C is repeatedly embedded (eight times), an indeterminate code can be identified by a majority decision at the time of decoding, and reliability is improved. Can be increased.

  Further, according to the first embodiment, the processing required for decoding the code for the image data in the decoder 200 can be reduced.

(Embodiment 2)
In the first embodiment described above, the configuration example in which the encoder 100 encodes the code C (see FIG. 4) itself and the decoder 200 decodes the code has been described. However, the code is converted into an error correction code (for example, BCH). An error correction code code encoded using (Bose Chaudhuri Hocquenghem) code, Reed-Solomon code, or the like) may be encoded and decoded on the decoder side. Hereinafter, this configuration example will be described as a second embodiment.

  In the following, as a specific example of the image data processing apparatus according to the second embodiment, an encoder 300 (see FIG. 12) for embedding a code (binary number) as another data in original image data and generating image coded data. And the decoder 400 (see FIG. 13) that reads the code from the image coded data will be described.

  FIG. 12 is a block diagram showing a configuration of the encoder 300 according to the second embodiment of the present invention. In this figure, parts corresponding to those in FIG. 1 are denoted by the same reference numerals, and description thereof is omitted.

  In FIG. 12, an error correction code encoding unit 301 is newly provided, and a comparison unit 302 and an encoding unit 303 are provided instead of the comparison unit 105 and the encoding unit 106 shown in FIG.

  The error correction code encoding unit 301 encodes a code Ca composed of x bits using an error correction code (for example, a BCH code), and generates an error correction code code CE.

  The error correction code CE includes a code Ca (x bits) and redundant bits (y bits) calculated from the code Ca and a predetermined calculation formula, and (x + y) bits (for example, 15 bits). ). According to this error correction code CE, for example, even if a 3-bit error occurs, the original code Ca can be restored by error correction.

  Here, the encoder 300 is an apparatus for embedding the error correction code CE in the original image data I0 (see FIG. 2), for example, eight times to generate the image coded data I4. The difference from the encoder 100 (see FIG. 1) is that an error correction code CE is used.

The comparison unit 302 is a bit determined from the n-th bit of the error correction code code CE and the magnitude relationship between the left average density data D _l and the right average density data D _r stored in the register 104 _l and the register 104 _r. The determination result (“0” or “1” is bit-determined by the relational expression described in the first embodiment: see FIG. 3) is compared.

Based on the comparison result of the comparison unit 302, the encoding unit 303 executes a process for embedding the error correction code code CE in the block divided image data I1 (original image data I0). Specifically, the encoding unit 303 maintains the magnitude relationship between the left average density data D _l and the right average density data D _r when the comparison results of the comparison unit 302 match, while the comparison results do not match. In some cases, the left average density data D ₁ and the right average density data _Dr are changed (the magnitude relation is reversed) so as to have a magnitude relationship representing the bits of the error correction code CE, thereby generating image coded data I4. And output this.

  The image coded data I4 has the same configuration as the image coded data I3 shown in FIG. Therefore, the same error correction code code CE (15 (x + y) bits) is embedded eight times in total in the image coded data I4.

  Each component of the encoder 300 is interconnected via a control unit (not shown).

  The operation of the encoder 300 is the same as the operation of the encoder 100 described above except that the error correction code code CE is used, and a description thereof will be omitted.

  FIG. 13 is a block diagram showing a configuration of the decoder 400 according to the second embodiment of the present invention. The decoder 400 is a device for decoding the embedded error correction code CE (code Ca) from the image encoded data I4 encoded by the encoder 300 (see FIG. 12).

  In this figure, the same reference numerals are given to the portions corresponding to the respective portions in FIG. In FIG. 13, a comparison unit 401 and a decoding unit 402 are provided instead of the comparison unit 206 and the decoding unit 207 shown in FIG.

Similarly to the comparison unit 206 (see FIG. 9), the comparison unit 401 compares the left average density data and the right average density data stored in the register 205 _l and the register 205 _r , thereby determining the bit. And a candidate error correcting code group CEG (candidate error correcting code CE ₁ corresponding to the bit determination result (determined as “0” or “1” by the relational expression described in the first embodiment)). To CE ₈ (see FIG. 15) are output to the decoding unit 402.

Here, each of the candidate error correction code codes CE _{1 to} CE ₈ shown in FIG. 15 has a 15 (x + y) bit configuration, and eight error correction code codes embedded in the image encoded data I4. (15 bits) is the result of decoding, and is a candidate for the error correction code code CE ′ (see FIG. 15).

In the candidate error correction code CE _{1 to} CE ₈ , “2” represents a bit for which the bit determination of “1” or “0” is indeterminate.

As shown in FIG. 15, the decoding unit 402 takes a majority decision in bit units (bit string in the vertical direction in the figure) from the candidate error correction code codes CE _{1 to} CE ₈ corresponding to the comparison result of the comparing unit 401. Bits (all 15 bits) are determined and set as an error correction code CE ′.

Strictly speaking, there is a possibility that an error bit is included in the error correcting code CE ′ by majority vote. Therefore, in the second embodiment, decoding section 402 uses “2” (indeterminate) bits for error correcting code in candidate error correcting code codes CE _{1 to} CE ₈ (candidate error correcting code group CEG). Complementary processing of replacing with the corresponding bit of the code CE ′ is executed to obtain candidate error correcting code codes CE ′ _{1 to} CE ′ ₈ (candidate error correcting code group CEG ′).

Further, the decoding unit 402 performs error correction code decoding processing (for example, decoding processing based on the BCH code) for each of the candidate error correction code codes CE ′ _{1 to} CE ′ ₈ , and codes (shown in FIG. 12). Corresponding to the code Ca) and the number of errors. The number of errors is the number of error bits in each of the error correction code codes CE ′ _{1 to} CE ′ ₈ .

  Further, the decoding unit 402 outputs, as a code Ca ′, a code corresponding to the minimum number of errors among the codes obtained by the error correction code decoding process. This code Ca ′ is a decoding result of the decoder 400.

  Each component of the decoder 400 is interconnected via a control unit (not shown).

  Next, the operation of the decoder 400 shown in FIG. 13 will be described with reference to the flowchart shown in FIG.

  In the figure, in step SD1, image coded data I4 (see FIGS. 12 and 13) is input to the image cutout unit 201. In step SD2, n is set to 1 as initialization. This n represents a bit pointer of the code to be decoded. In this case, n = 1 corresponds to the leftmost bit of the code.

  In step SD3, when the image cutout unit 201 includes image data (for example, a blank portion) around the input image coded data I4, the image cutout unit 201 cuts out valid image coded data I4 from the whole.

  In step SD4, the block dividing unit 202 divides the image coded data I4 cut out by the image cutout unit 201 into M × N blocks by block division processing, and this is divided into block divided image data (not shown). The data is output to the block extraction unit 203.

  In step SD5, the block extraction unit 203 extracts a pair block (two blocks) corresponding to n = 1 from block divided image data (not shown), and then averages the density distribution in each block as block density data. Output to the unit 204.

  In step SD6, the averaging unit 204 performs an averaging process to calculate left average density data (not shown) corresponding to one block and right average density data (not shown) corresponding to the other block from the block density data. And ask.

At step SD7, the averaging unit 204, the left-side average density data (not shown) in the register 205 _l, respectively store the right average density data (not shown) in the register 205 _r.

In step SD8, comparing unit 401, by comparing the left average density data and the magnitude relation between the right average density data stored in the registers 205 _l and the register 205 _r, performs bit decision, the bit determination result (the above-mentioned The bit is determined as “0” or “1” according to the relational expression).

  Here, the comparison unit 401 obtains a density difference between the left average density data and the right average density data, and when this density difference is larger than the upper limit threshold value described in step SA14 (see FIG. 6), Assuming that the bit determination result corresponding to the pair block is unreliable, the bit determination result is “2” (indeterminate: see FIG. 15).

  In step SD9, the comparison unit 401 increments n by 1. As a result, n is set to 2. In step SD10, the comparison unit 401 determines whether n is larger than nend (= 15) in the same manner as in step SC10 (see FIG. 11). In this case, the determination result is “No”. .

  Thereafter, the operations after step SD5 are repeated until the determination result of step SD10 is “Yes”.

When n is set to 17 in step SD9, the determination result in step SD10 is “Yes”. At this point, the comparison unit 401 uses the candidate error correcting code code CE ₁ shown in FIG. 15 as the bit discrimination result.

  In step SD11, the comparison unit 401 determines whether or not the above-described processing relating to the last pair block (two blocks) in the block-divided image data (not shown) has been completed. In this case, the determination result is “No”. And

  In step SD16, the comparison unit 401 sets n to 1 to reset n (= 16). In step SD5, the block extraction unit 203 extracts the next pair block (two blocks) corresponding to n = 1 from the block divided image data (not shown), and then uses the density distribution in each block as block density data. The result is output to the averaging unit 204.

  Thereafter, the above-described operation is repeated until the determination result in step SD11 becomes “Yes”.

When the determination result in step SD11 is “Yes”, in step SD12, the decoding unit 402 executes majority decision processing in the same manner as in step SC12 (see FIG. 11). That is, at this time, the candidate error correcting code codes CE _{1 to} CE ₈ shown in FIG. 15 are the bit determination results.

The decoding unit 402 takes a majority vote in bit units (bit string in the vertical direction in the figure) from the candidate error correction code codes CE _{1 to} CE ₈ and determines each bit (15 bits in total). As a result, an error correcting code CE ′ is generated.

For example, in the case of the leftmost bits of the candidate error correcting code CE _{1 to} CE ₈ , “0” is 2, “1” is 5, and “2” is 1, so that the error correcting code CE ′ The leftmost bit of is determined as “1” by majority vote.

In step SD13, the decoding unit 402 converts the bit of “2” (indeterminate) into the error correction code code CE ′ in the candidate error correction code codes CE _{1 to} CE ₈ (candidate error correction code group CEG). A complementary process of replacing with the corresponding bit is executed to obtain a candidate error correcting code group CEG ′ (candidate error correcting code codes CE ′ _{1 to} CE ′ ₈ ).

In step SD14, the decoding unit 402 performs error correction code decoding processing (for example, decoding processing based on the BCH code) for each of the candidate error correction code codes CE ′ _{1 to} CE ′ ₈ by error correction code decoding processing. The code (corresponding to the code Ca shown in FIG. 12) and the number of errors are obtained.

  In step SD15, the decoding unit 402 outputs, as a code Ca ', a code corresponding to the minimum number of errors among the codes obtained by the error correction code decoding process.

  As described above, according to the second embodiment, as described with reference to FIG. 12, since the error correction code CE is embedded in a plurality of blocks, error correction in the decoder 400 or the like is possible. Thus, reliability can be improved.

(Embodiment 3)
In the first and second embodiments described above, as illustrated in FIG. 3, the configuration example in which encoding and decoding are performed using the average density of the entire block (for example, the block B ₁₁₈ ) has been described. It is good also as a structural example using the average density | concentration of a part (for example, center part) of a block. Hereinafter, this configuration example will be described as a third embodiment.

FIG. 16 is a diagram for explaining the third embodiment according to the present invention. In the figure, a block B corresponds to the block (for example, the block B ₁₁₈ shown in FIG. 3) in the first and second embodiments, and has a t × t pixel size. In the third embodiment, encoding and decoding are performed using the average density in the central portion Ba ((t / 2) × (t / 2) pixel size) as a part of the block B.

  As described above, according to the third embodiment, it is possible to reduce the processing amount for the density changing process compared to the case where the average density of the entire block is used.

(Embodiment 4)
When the first to third embodiments described above are applied to a color image, a method of embedding a code in a grayscale image (monochrome image) obtained by converting a color image into lightness, and three primary colors (cyan, magenta, A method of embedding codes in any one of the yellow color components (cyan component, magenta component, yellow component) (for example, yellow component) is conceivable.

  Compared to the former method, the latter method has an advantage that it is difficult for human eyes to distinguish even if the density is changed (see FIG. 18). FIG. 18 is a graph showing the correspondence between the concentration change amount and MOS (Mean Opinion Score) in the fourth embodiment.

  This graph is a result of evaluation by a subjective evaluation method defined by the International Lighting Association (ITU-T), which is well known as an image evaluation method. In the subjective evaluation method, when the apparent difference between the original image and the post-change image that has been changed by the amount of density change is not known, the MOS value is evaluated as 5 points, and the MOS value is decreased as the difference increases. The

  In the graph, “yellow” is an evaluation result regarding the image of the yellow component. “Brightness” is an evaluation result regarding a grayscale image. As can be seen from this graph, the yellow component (“yellow”) has a higher MOS value than the gray scale (“lightness”) even when the density change amount is increased. Therefore, the latter method of changing the density of the yellow component is difficult to distinguish with the human eye.

  Hereinafter, a configuration example in which a code is embedded in a color component (in this case, a yellow component) will be described as a fourth embodiment.

  FIG. 17 is a block diagram showing a configuration of an encoder 500 according to the fourth embodiment of the present invention. In this figure, the same reference numerals are given to portions corresponding to the respective portions in FIG. 12, and description thereof is omitted.

  In FIG. 17, a yellow component cutout unit 501 and a yellow component integration unit 502 are newly provided. The yellow component cutout unit 501 cuts out a yellow component from the block density data D, and outputs this to the averaging unit 103 as yellow block density data DE.

Further, the yellow component cutout unit 501 obtains the left average density data DE _l corresponding to one block and the right average density data DE _r corresponding to the other block from the yellow block density data DE, and registers them. 104 _l and register 104 _r .

  Further, the yellow component cutout unit 501 cuts out the cyan component and the magenta component from the block density data D, and outputs them to the yellow component integration unit 502 as cyan / magenta block density data DCM.

  In the fourth embodiment, the process is executed on the yellow block density data DE, and the density of the yellow component is changed. The encoding unit 303 outputs yellow image coded data IE4 corresponding to the yellow component. The yellow component integration unit 502 integrates the yellow image coded data IE4 and the cyan / magenta block density data DCM and outputs the result as image coded data I5.

  As described above, according to the fourth embodiment, since the yellow component is cut out by the yellow component cutout unit 501 and embedded in a plurality of blocks of the yellow component, the characteristic that yellow is not conspicuous is obtained. By utilizing this, it is possible to maintain the data discrimination capability without degrading the image quality.

(Embodiment 5)
In the first embodiment, the image clipping in the image clipping unit 201 illustrated in FIG. 9 has not been described in detail. However, as illustrated in FIG. 19, the margins P are formed at the four corners of the image encoded data I3. the yellow mark ME ₁ ~ME ₄ for the distinction may be configured example of adding the encoder 100 (see FIG. 1). Hereinafter, this configuration example will be described as a fifth embodiment.

In FIG. 19, yellow marks ME _{1 to} ME ₄ are yellow micro dots, and are hardly noticeable as described with reference to FIG.

Next, image clipping in the image clipping unit 201 illustrated in FIG. 9 will be described. When the margin P and the image coded data I3 (yellow marks ME _{1 to} ME ₄ ) shown in FIG. 19 are input to the image cutout unit 201, the four corner search process shown in FIG. 21 is executed.

In this four-corner search process, four corners (yellow marks ME _{1 to} ME ₄ ) of the image coded data I3 shown in FIG. 19 are searched. Hereinafter, the pixel value of the margin P (white) is set to 255, and the pixel value of black (yellow) is set to 0.

  Specifically, in step SE1 shown in FIG. 20, the image cutout unit 201 searches for the minimum pixel value (for example, 250) in the blank portion P (white) around the image encoded data I3. In step SE2, the image cutout unit 201 sets a threshold value (225) by multiplying the minimum pixel value (250) by the safety factor (0.9).

  In step SE3, as shown in FIG. 21, the image cut-out unit 201 moves the search line L at an angle of 45 degrees in the X direction from the corner of the margin P (upper left corner in the figure) while moving the threshold value. (255) The following pixel positions are searched.

When the search line L reaches the yellow mark ME _1, in the pixel value of the yellow marks ME ₁ (= 0) because it is below the threshold, the image cut-out unit 201, the four corners of the image coded data I3 As a corner, the position of the yellow mark ME ₁ is determined.

In step SE4, the image cutout unit 201 determines whether or not the four corners of the image encoded data I3 have been searched. In this case, the determination result is “No”. Thereafter, the process of step SE3 is executed for the other three corners. When the four corners (yellow marks ME _{1 to} ME ₄ ) shown in FIG. 19 are searched, the image cutout unit 201 sets “Yes” as a result of the determination made at step SE4.

Next, the image cutout unit 201 uses the position coordinates of the yellow marks ME _{1 to} ME ₄ to perform a known affine transformation or the like from the entire image (margin portion P and image coded data I3) shown in FIG. The image coded data I3 is cut out.

As described above, according to the fifth embodiment, as shown in FIG. 19, the yellow marks ME _{1 to} ME ₄ for clipping are added to the image encoded data I3 that is the embedding result. It is possible to easily cut out an image at the time of decoding.

  Although Embodiments 1 to 5 according to the present invention have been described in detail with reference to the drawings, specific configuration examples are not limited to these Embodiments 1 to 5 and depart from the gist of the present invention. Even if there is a design change or the like within a range not to be included, it is included in the present invention.

  For example, in Embodiments 1 to 5 described above, a program for realizing the functions of the encoder 100, the decoder 200, the encoder 300, the decoder 400, or the encoder 500 is recorded in the computer-readable recording medium 700 shown in FIG. Then, each function may be realized by causing the computer 600 shown in the figure to read and execute the program recorded in the recording medium 700.

  A computer 600 shown in the figure includes a CPU (Central Processing Unit) 610 that executes the above-described program, an input device 620 such as a keyboard and a mouse, a ROM (Read Only Memory) 630 that stores various data, an operation parameter, and the like. RAM (Random Access Memory) 640, a reading device 650 for reading a program from the recording medium 700, and an output device 660 such as a display and a printer.

  The CPU 610 implements the above-described functions by reading a program recorded on the recording medium 700 via the reading device 650 and then executing the program. Examples of the recording medium 700 include an optical disk, a flexible disk, and a hard disk.

  Further, in the first to fifth embodiments, the example in which the average density is used as the feature amount of the image (block) has been described. However, the present invention is not limited to this, and the granularity, saturation, density gravity center, dispersion, and the like are not limited thereto. Other feature amounts obtained from the image may be used. In this case, since the feature amount is granularity, saturation, density gravity center or dispersion, it is difficult to notice and the influence of image quality can be reduced.

  In the first to fifth embodiments, the feature amount converted into another feature amount based on a predetermined conversion rule may be used.

(Supplementary Note 1) Dividing means for dividing image data into a plurality of blocks;
Feature quantity extraction means for extracting each feature quantity in the plurality of blocks;
Encoding means for associating one code with each pair block in the plurality of blocks based on the magnitude relationship of the feature amount, and embedding the plurality of codes in the plurality of blocks;
An image data processing apparatus comprising:

(Supplementary note 2) The encoding means, when the magnitude relation of the feature quantity in the pair block does not match the code to be embedded, changes the feature quantity in the pair block so as to reverse the magnitude relation. The image data processing apparatus according to appendix 1.

(Additional remark 3) The said encoding means does not change the said feature-value, when the difference of the feature-value between the blocks in the said pair block exceeds an upper limit threshold value, The image data of Additional remark 2 characterized by the above-mentioned Processing equipment.

(Supplementary note 4) When the encoding unit matches the code of the feature value in the pair block with the code to be embedded, and the difference in the feature value between the blocks in the pair block is less than a lower threshold, 4. The image data processing apparatus according to any one of appendices 1 to 3, wherein the feature amount in the pair block is changed so that the difference is equal to or greater than a lower limit threshold value.

(Supplementary Note 5) Dividing means for dividing image data into a plurality of blocks;
Error correction code code generation means for encoding a code using an error correction code and generating a code for the error correction code;
Feature quantity extraction means for extracting each feature quantity in the plurality of blocks;
Encoding means for associating a code for one error correction code with each pair block in the plurality of blocks based on the magnitude relationship of the feature quantity, and embedding a plurality of codes for error correction code in the plurality of blocks;
An image data processing apparatus comprising:

(Supplementary note 6) The image data processing device according to any one of supplementary notes 1 to 5, wherein the feature amount extraction unit extracts each feature amount in a part of each block.

(Supplementary note 7) Any one of Supplementary notes 1 to 6, further comprising: a cutout unit that cuts out a yellow component from the plurality of blocks, wherein the encoding unit embeds the plurality of blocks of the yellow component. The image data processing device described in 1.

(Supplementary note 8) The image data processing apparatus according to any one of supplementary notes 1 to 7, wherein the encoding unit adds a cutout mark pixel for cutout to an image that is an embedding result.

(Supplementary note 9) The image data processing apparatus according to any one of supplementary notes 1 to 8, wherein the feature amount is an average density.

(Supplementary note 10) The image data processing apparatus according to any one of supplementary notes 1 to 8, wherein the feature amount is granularity, saturation, density gravity center, or dispersion.

(Supplementary note 11) The image data processing apparatus according to any one of supplementary notes 1 to 10, wherein the encoding unit repeatedly embeds a plurality of codes.

(Supplementary note 12) An image data processing device that decodes the plurality of codes from an image that is an embedding result in the image data processing device according to any one of Supplementary notes 1 to 11.

(Supplementary note 13) a dividing step of dividing the image data into a plurality of blocks;
A feature amount extraction step of extracting each feature amount in the plurality of blocks;
An encoding step of associating one code with each pair block in the plurality of blocks based on the magnitude relation of the feature amount, and embedding the plurality of codes in the plurality of blocks,
An image data processing method comprising:

(Additional remark 14) In the said encoding process, when the magnitude relationship of the said feature-value in the said pair block does not correspond with the code which should be embedded, the said feature-value in the said pair block is changed so that a magnitude relationship may be reversed, It is characterized by the above-mentioned. The image data processing method according to appendix 13.

(Supplementary note 15) The image data according to supplementary note 14, wherein, in the encoding step, the feature quantity is not changed when a difference in feature quantity between the blocks in the pair block exceeds an upper limit threshold value. Processing method.

(Supplementary Note 16) In the encoding step, when the magnitude relationship of the feature amount in the pair block matches a code to be embedded, and the difference in feature amount between the blocks in the pair block is less than a lower threshold, The image data processing method according to any one of appendices 13 to 15, wherein the feature amount in the pair block is changed so that the difference is equal to or greater than a lower threshold value.

(Supplementary note 17) a division step of dividing image data into a plurality of blocks;
An error correction code code generation step for encoding the code using an error correction code and generating a code for the error correction code;
A feature amount extraction step of extracting each feature amount in the plurality of blocks;
An encoding step of associating a code for one error correction code with each pair block in the plurality of blocks based on the magnitude relation of the feature amount, and embedding a plurality of codes for the error correction code in the plurality of blocks;
An image data processing method comprising:

(Supplementary note 18) The image data processing method according to any one of supplementary notes 13 to 17, wherein in the feature quantity extraction step, each feature quantity in a part of each block is extracted.

(Supplementary note 19) Any one of Supplementary notes 13 to 18, including a cutting step of cutting out a yellow component from the plurality of blocks, wherein the encoding step includes embedding a plurality of blocks of the yellow component. The image data processing method described in 1.

(Supplementary note 20) The image data processing method according to any one of supplementary notes 13 to 19, wherein in the encoding step, a cutout mark pixel for cutout is added to an image that is an embedding result.

(Supplementary note 21) The image data processing method according to any one of supplementary notes 13 to 20, wherein the feature amount is an average density.

(Supplementary note 22) The image data processing method according to any one of supplementary notes 13 to 20, wherein the feature amount is granularity, saturation, density gravity center, or dispersion.

(Supplementary note 23) The image data processing method according to any one of supplementary notes 13 to 22, wherein in the encoding step, a plurality of codes are repeatedly embedded.

(Supplementary Note 24) An image data processing method for decoding the plurality of codes from an image that is an embedding result in the image data processing method according to any one of Supplementary notes 13 to 23.

(Appendix 25) Computer
A dividing means for dividing the image data into a plurality of blocks;
Feature quantity extraction means for extracting each feature quantity in the plurality of blocks;
Encoding means for associating one code with each pair block in the plurality of blocks based on the magnitude relation of the feature amount, and embedding the plurality of codes in the plurality of blocks;
Image data processing program for functioning as

(Supplementary note 26) When the magnitude relationship of the feature values in the pair block does not match the code to be embedded, the encoding unit changes the feature value in the pair block so as to reverse the magnitude relationship. The image data processing program according to attachment 25.

(Supplementary note 27) The image data according to supplementary note 26, wherein the encoding unit does not change the feature amount when a difference in feature amount between the blocks in the pair block exceeds an upper limit threshold value. Processing program.

(Supplementary note 28) When the encoding unit matches the code of the feature amount in the pair block with a code to be embedded, and the difference in feature amount between the blocks in the pair block is less than a lower threshold, The image data processing program according to any one of appendices 25 to 27, wherein the feature amount in the pair block is changed so that the difference is equal to or greater than a lower limit threshold value.

(Supplementary note 29)
A dividing means for dividing the image data into a plurality of blocks;
Error correction code code generation means for encoding a code using an error correction code and generating a code for the error correction code;
Feature quantity extraction means for extracting each feature quantity in the plurality of blocks;
Encoding means for associating one error correction code code with each pair block in the plurality of blocks based on the magnitude relationship of the feature quantity, and embedding the plurality of error correction code codes in the plurality of blocks,
Image data processing program for functioning as

(Supplementary Note 30) The image data processing program according to any one of Supplementary Notes 25 to 29, wherein the feature amount extraction unit extracts each feature amount in a part of each block.

(Supplementary note 31) The computer is caused to function as a cutting unit that cuts out a yellow component from the plurality of blocks, and the encoding unit embeds the plurality of blocks of the yellow component. 30. The image data processing program according to any one of 30.

(Supplementary note 32) The image data processing program according to any one of supplementary notes 25 to 31, wherein the encoding unit adds a cutout mark pixel for cutout to an image that is an embedding result.

(Supplementary note 33) The image data processing program according to any one of supplementary notes 25 to 32, wherein the feature amount is an average density.

(Supplementary note 34) The image data processing program according to any one of supplementary notes 25 to 32, wherein the feature amount is granularity, saturation, density gravity center, or dispersion.

(Supplementary Note 35) The image data processing program according to any one of Supplementary Notes 25 to 34, wherein the encoding means repeatedly embeds a plurality of codes.

(Supplementary note 36) An image data processing program for decoding the plurality of codes from an image as an embedding result in the image data processing program according to any one of supplementary notes 25 to 35.

  As described above, the image data processing device, the image data processing method, and the image data processing program according to the present invention are useful when data is embedded in an image and used particularly when data is embedded in a print image. Suitable for cases.

It is a block diagram which shows the structure of the encoder 100 in Embodiment 1 concerning this invention. It is a figure which shows the original image data I0 shown in FIG. It is a figure which shows the block division | segmentation image data I1 shown in FIG. It is a figure which shows the code | cord | chord C shown in FIG. It is a figure which shows the image coding data I3 shown in FIG. 3 is a flowchart illustrating an operation example 1 of the encoder 100 illustrated in FIG. 1. It is a figure explaining the density change process in the encoder 100 shown in FIG. 6 is a flowchart illustrating an operation example 2 of the encoder 100 illustrated in FIG. 1. 2 is a block diagram showing a configuration of a decoder 200 in the first embodiment. FIG. It is a figure explaining the majority determination process in the decoder 200 shown in FIG. 10 is a flowchart for explaining the operation of the decoder 200 shown in FIG. 9. It is a block diagram which shows the structure of the encoder 300 in Embodiment 2 concerning this invention. It is a block diagram which shows the structure of the decoder 400 in the same Embodiment 2. 14 is a flowchart for explaining the operation of the decoder 400 shown in FIG. 13. It is a figure explaining the code determination process in the decoder 400 shown in FIG. It is a figure explaining Embodiment 3 concerning this invention. It is a block diagram which shows the structure of the encoder 500 in Embodiment 4 concerning this invention. 10 is a graph showing a relationship between a density change amount and a MOS value in the fourth embodiment. It is a figure explaining the image cutting-out process in Embodiment 5 concerning this invention. 16 is a flowchart for describing four corner search processing according to the fifth embodiment. It is a figure explaining the four corner search process in the same Embodiment 5. FIG. It is a block diagram which shows the structure of the modification of Embodiment 1-5 concerning this invention.

Explanation of symbols

100 encoder 101 block dividing unit 102 block extracting unit 103 averaging unit 104 _l, 104 _r register 105 comparator unit 106 encoding unit 200 decoder 201 image cutting unit 202 block dividing unit 203 block extracting unit 204 averaging unit 205 _l, 205 _r Register 206 Comparison unit 207 Decoding unit 300 Encoder 301 Error correction code encoding unit 302 Comparison unit 303 Encoding unit 400 Decoder 401 Comparison unit 402 Decoding unit 500 Encoder 501 Yellow component extraction unit 502 Yellow component integration unit

Claims (27)

In an image data processing apparatus for extracting a code embedded in image data,
A dividing means for dividing the image data into a plurality of blocks;
Feature quantity extraction means for extracting feature quantities in each block of the plurality of blocks;
Decoding means for extracting a code from the pair block based on the magnitude relationship of the feature amount between each block of a pair block consisting of two blocks;
With
The decoding means includes
When the feature amount difference between the blocks in the pair block is not equal to or greater than the upper threshold value and the feature amount of the right block in the pair block is larger than the feature amount of the left block, the bit determination result from the pair block Extract the first code,
When the difference in feature quantity between the blocks in the pair block is not equal to or greater than the upper threshold value and the feature quantity in the left block in the pair block is larger than the feature quantity in the right block, the bit determination result from the pair block Extract the second code,
If the difference in the feature amount between the blocks in the pair block is equal to or greater than an upper threshold, the third code is extracted from the pair block as an uncertain bit determination result,
An image data processing apparatus that extracts a code based on the first code, the second code, and the third code.   2. The image data processing apparatus according to claim 1, further comprising error correction processing means for correcting an error of a code embedded in an image using the code obtained by the decoding means.   The image data processing apparatus according to claim 1, wherein the feature amount extraction unit extracts a feature amount in a part of each block.   The image data processing apparatus according to claim 3, wherein the feature amount extraction unit extracts a feature amount in a central portion of each block. The feature amount extraction unit extracts a feature amount of a specific color component included in image data,
5. The code according to claim 1, wherein the decoding unit extracts a code embedded in a plurality of blocks based on the feature amount of the specific color component extracted by the feature amount extraction unit. The image data processing apparatus described.   The image data processing apparatus according to claim 5, wherein the specific color component is a yellow component.   The image data processing apparatus according to claim 1, wherein the pair block is two adjacent blocks.   The image data processing apparatus according to claim 1, wherein the dividing unit divides the image based on cutout cutout mark pixels added to the image.   The image data processing apparatus according to claim 1, wherein the feature amount is at least one of average density, granularity, saturation, density gravity center, and variance. In an image data processing method for extracting a code embedded in image data,
A division step of dividing the image data into a plurality of blocks;
A feature amount extraction step of extracting a feature amount in each block of the plurality of blocks;
A decoding step of extracting a code from the pair block based on the magnitude relationship of the feature quantity between each block of a pair block consisting of two blocks;
Including
The decoding step includes
When the feature amount difference between blocks in the pair block is not equal to or greater than the upper threshold value and the feature amount of the right block in the pair block is larger than the feature amount of the left block, the bit determination result from the pair block Extract the first code,
When the difference in the feature amount between the blocks in the pair block is not equal to or greater than the upper threshold value and the feature amount in the left block in the pair block is larger than the feature amount in the right block, the bit determination result from the pair block Extract the second code,
If the difference in feature amount between the blocks in the pair block is equal to or greater than an upper threshold, the third code is extracted from the pair block as an uncertain bit determination result,
An image data processing method, comprising: extracting a code based on the first code, the second code, and the third code.   The image data processing method according to claim 10, further comprising an error correction processing step of correcting an error of a code embedded in an image using the code obtained by the decoding step.   12. The image data processing method according to claim 10, wherein the feature amount extraction step extracts a feature amount in a part of each block.   The image data processing method according to claim 12, wherein the feature amount extraction step extracts a feature amount in a central portion of each block. The feature amount extraction step extracts a feature amount of a specific color component included in the image data,
14. The code according to claim 10, wherein the decoding step extracts a code embedded in a plurality of blocks based on a feature amount of the specific color component extracted by the feature amount extraction step. The image data processing method described.   The image data processing method according to claim 14, wherein the specific color component is a yellow component.   16. The image data processing method according to claim 10, wherein the pair block is two adjacent blocks.   The image data processing method according to any one of claims 10 to 16, wherein in the dividing step, the image is divided based on cut-out mark pixels for clipping given to the image.   The image data processing method according to claim 10, wherein the feature amount is at least one of average density, granularity, saturation, density gravity center, and variance. In an image data processing program for extracting a code embedded in image data,
Computer
A dividing means for dividing the image data into a plurality of blocks;
Feature quantity extraction means for extracting feature quantities in each of the plurality of blocks;
Decoding means for extracting a code from the pair block based on the magnitude relationship of the feature quantity between the blocks of the pair block consisting of two blocks;
Function as
The decoding means includes
When the feature amount difference between the blocks in the pair block is not equal to or greater than the upper threshold value and the feature amount of the right block in the pair block is larger than the feature amount of the left block, the bit determination result from the pair block Extract the first code,
When the difference in feature quantity between the blocks in the pair block is not equal to or greater than the upper threshold value and the feature quantity in the left block in the pair block is larger than the feature quantity in the right block, the bit determination result from the pair block Extract the second code,
If the difference in the feature amount between the blocks in the pair block is equal to or greater than an upper threshold, the third code is extracted from the pair block as an uncertain bit determination result,
An image data processing program that extracts a code based on the first code, the second code, and the third code. The computer,
20. The image data processing program according to claim 19, wherein the image data processing program further functions as error correction processing means for correcting an error of a code embedded in an image using the code obtained by the decoding means.   21. The image data processing program according to claim 19, wherein the feature amount extraction unit extracts a feature amount in a part of each block.   The image data processing program according to claim 21, wherein the feature amount extraction unit extracts a feature amount in a central portion of each block. The feature amount extraction unit extracts a feature amount of a specific color component included in image data,
23. The code according to claim 19, wherein the decoding unit extracts a code embedded in a plurality of blocks based on the feature amount of the specific color component extracted by the feature amount extraction unit. The image data processing program described.   The image data processing program according to claim 23, wherein the specific color component is a yellow component.   The image data processing program according to any one of claims 19 to 24, wherein the pair block is two adjacent blocks.   The image data processing program according to any one of claims 19 to 25, wherein the dividing unit divides the image on the basis of the cut-out mark pixel for cutting added to the image.   The image data processing program according to any one of claims 19 to 26, wherein the feature amount is at least one of average density, granularity, saturation, density gravity center, and variance.

Images