JP2010011438A - Image processing apparatus and method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image processing apparatus and method that embeds information in an image having been already subjected to gradation processing and even an image having been subjected to gradation processing smaller in tone number. <P>SOLUTION: The image processing apparatus includes a gradation processing means (105) of performing gradation processing on an image, a signal generating means (101) of generating a data signal representing information different from the image, and an information embedding means (102, 106) of embedding the different information in the image by adding a striped pattern having a plurality of frequency components corresponding to the data signal to the image signal after the gradation processing. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a technique for embedding information in a document image or the like.

  Recently, digital watermarking is known as one of effective techniques used for copyright protection and security purposes. This digital watermark technique embeds data other than the content data in the content data. For example, for content such as audio and video, information such as information specifying the copyright holder, information indicating whether copying is possible, etc. In a still image, information related to the image and information such as copyright are embedded. Whether or not the embedded information can be read illegally can be determined based on whether or not the embedded information can be read from the content.

  By the way, as a method of embedding in a still image, a method of embedding information by exchanging lower bit values of pixel values is known. However, in this method, the fluctuation of the pixel value accompanying information embedding is very small, so if the still image is duplicated with a hard copy with large image quality degradation, the embedded information can be read from the hard copy. There is no problem.

  For this reason, a method for frequency analysis of an output image obtained by embedding information in content data by using a periodic component, or a method of embedding by modulating a pattern at the time of gradation processing is disclosed (for example, , See Patent Document 1).

  However, all of these conventional information embedding techniques make fine changes to the image data before gradation processing, and then perform gradation processing. The reason for this is that if data is embedded by a conventional method after gradation processing, the gradation pattern changes and timing cannot be obtained. As described above, when the image processing system is already configured as one, even if a function related to data embedding processing is added later, it is difficult to redesign the entire image processing system. There were many things.

  In the embedding process, since the number of gradations of the original image is changed, information may not be embedded if the number of gradations of the original image is small.

  Therefore, information embedding processing functions can be easily added to existing image processing systems in the trend of enhanced copyright protection and security enhancement, and information embedding processing can be applied to images with a small number of gradations. Is desired.

  The present invention has been made in view of such circumstances, and an image processing apparatus capable of embedding information in an image that has already been subjected to gradation processing, and an image that has been subjected to gradation processing with a small number of gradations, and An object is to provide an image processing method.

  In order to solve the above-described problems, the present invention provides a gradation processing unit that performs gradation processing on an image, a signal generation unit that generates a data signal that represents information different from the image, and a response to the data signal. And an information embedding unit that embeds the other information in the image by adding a stripe pattern having a plurality of frequency components to the image signal after the gradation processing.

  The present invention also performs gradation processing on an image, generates a data signal representing information different from the image, and applies a stripe pattern having a plurality of frequency components corresponding to the data signal after the gradation processing. The image processing method embeds the other information in the image by adding to the image signal.

  According to the present invention, it is possible to provide an image processing apparatus and an image processing method capable of embedding information in an image that has already been subjected to gradation processing and an image that has been subjected to gradation processing with a small number of gradations. .

1 is a block diagram showing an image processing apparatus according to a first embodiment of the present invention. The figure explaining the embedding principle of a watermark pattern. The figure which showed an example of arrangement | positioning on the frequency surface of the periodic component embedded in object image data. The figure which shows the point on the main scanning axis on a frequency surface, or near a main scanning axis. The block diagram which shows the image processing apparatus of the 2nd Embodiment of this invention. The block diagram which shows the image processing apparatus of the 3rd Embodiment of this invention. The figure showing an example of the operator for filtering used for the smoothing process. The figure which shows transition of the signal by which the image process was carried out. The block diagram which shows the image processing apparatus of the 4th Embodiment of this invention.

[First embodiment]
FIG. 1 is a block diagram showing an image processing apparatus according to the first embodiment of the present invention. The configuration and operation of the image processing apparatus will be described with reference to FIG.
The image processing apparatus 1 includes a data generation unit 101, a pattern generation unit 102, an image data input unit 103, an image processing unit 104, a gradation processing unit 105, and a pattern embedding unit 106.

  The data generation unit 101 generates data used for the watermark. The pattern generation unit 102 generates a watermark pattern to be embedded in the image from the generated data. The data (hereinafter referred to as “specific information”) provided from the data generation unit 101 is information set in advance or designated by the user, such as creation date and time, information related to an image to be embedded.

The image data input unit 103 inputs image data to be embedded from an external device (not shown), a reading device (not shown), or the like. The image processing unit 104 performs predetermined color conversion processing, γ conversion processing, inking processing, and the like on the input image data. The gradation processing unit 105 executes gradation processing. Here, it is desirable that the quantization level of the pseudo gradation processing used in the gradation processing unit 105 is a predetermined level, for example, about 16 values or more.
Note that the image processing unit 104 and the gradation processing unit 105 need not be provided separately in the order shown in FIG.

  The pattern embedding unit 106 embeds the watermark pattern output from the pattern generation unit 102 in the target image data output from the gradation processing unit 105. Then, the pattern embedding unit 106 supplies the image data in which the watermark pattern is embedded to the output / display unit 107. The output / display unit 107 outputs image data in which a watermark pattern is embedded.

FIG. 2 is a diagram for explaining the principle of embedding a watermark pattern.
The embedding of the watermark pattern is realized by superimposing a periodic pattern signal composed of a rectangular wave composed of ON (= 1) and OFF (= 0) corresponding to the specific information on the image data. The setting of the rectangular wave amplitude α and the rectangular wave period τ of the periodic pattern signal needs to be considered in consideration of human visual limitations. The smaller the amplitude and the shorter the period of the pattern, the less noticeable it is to the human eye.

  The pattern signal is superimposed so that the intermediate level of the amplitude becomes 0 level. Therefore, assuming that the amplitude in FIG. 2 is ± α / 2, the gradation value of ± α / 2 is superimposed on the image data on which the pattern signal is superimposed. The sign + indicates that the code bit representing the specific information is “1”, and the sign − indicates that the code bit is “0”.

  In the example shown in FIG. 2, a pattern obtained by amplitude-modulating embedded data at a constant period is superimposed on an image. However, as described above, in this method, there is a case where the embedded specific information cannot be read from a hard copy with a large image quality deterioration. Therefore, in the first embodiment, a large number of frequency components on the two-dimensional Fourier transform plane are multiplexed according to the specific information, and a two-dimensional striped pattern having multiple frequency components is added to the target image data. Embed with.

  FIG. 3 is a diagram showing an example of the arrangement on the frequency plane of the periodic components embedded in the target image data.

  In order to generate a striped pattern signal having multiple frequency components, the pattern generation unit 102 defines a Fourier transform plane composed of an axis in the main scanning direction and an axis in the sub scanning direction as shown in FIG. A large number of points arranged according to a predetermined rule are provided on the Fourier transform plane. Each bit data constituting the code of the specific information is arranged at a number of points according to a predetermined rule.

  For example, as indicated by numbers in the left area of FIG. 3, the bit data constituting the code of the specific information at each point in the order of 1 → 2 → 3 →... Starts from S (start point). Assigned.

  A distance f from the origin of the Fourier transform plane to each point represents a frequency (CPI: Cycle Per Inch). Here, since the period r corresponds to the reciprocal of the frequency f, it can be considered that each point on the Fourier transform plane represents the period r. On the other hand, the bit data assigned to each point represents the amplitude. Therefore, the period r is associated with the amplitude α by specifying each point.

  The specific information code generated by the data generation unit 101 is supplied to the pattern generation unit 102. When the code is supplied to the pattern generation unit 102, a plurality of bits constituting the code are sequentially arranged at predetermined positions on the Fourier transform plane. Note that the arrangement location and arrangement order of each bit can be arbitrarily determined.

  The points represented by black circles shown in FIG. 3 are basic periodic components used for pseudo gradation expression. The basic periodic component is, for example, an upper limit value (Nyquist frequency) of a high frequency component that can be expressed by the printer. When selecting each point, it is necessary to arrange each point so that the embedded periodic component does not cause an erroneous detection when reading embedded information due to interference with the basic periodic component.

  In other words, the arrangement of the embedded periodic components is set by combining angles and frequencies so that each point does not interfere with the black point as much as possible on the Fourier transform plane by avoiding the use of the periodic component of the same angle represented by a black circle. . It is necessary to embed information using a frequency that is the basis of gradation expression and its neighboring frequencies, specifically, points on the main scanning axis or avoiding points close to the main scanning axis as shown in FIG. is there.

  The pattern generation unit 102 generates an embedding pattern according to the position of the pixel in which the image data is embedded from the period r and the amplitude α of all the bit data of the specific information arranged on the Fourier transform plane.

The embedding pattern is generated by performing inverse Fourier transform on data defined on the Fourier transform plane. First, an addition value Δd is obtained according to the following equation (1).

  Here, αm is the amplitude of the periodic component to be embedded, r is the period (dot / cycle) of the periodic component, θm is the angle between the periodic component and the main scanning direction, x and y are the main scanning direction of the target pixel, and sub-scanning Represents the position of each direction.

Next, Δd obtained by adding all the periodic components is transformed into an added value Δd ′ using Equation (2) so as to match the multi-value number (quantization level) processed by the gradation processing unit 105.

  Here, int is a function (Gaussian symbol) that truncates the decimal value of the operation result and converts it to an integer value, and stp is converted so as to be adapted to the quantization multi-value number used in the gradation processing unit 105. Represents a conversion factor.

  The added value Δd obtained by Equation (1) is a value corresponding to the signal before gradation conversion. Therefore, by multiplying by the conversion coefficient stp, the value before gradation processing is converted to correspond to the value after gradation processing. Further, the reason why 0.5 is added to the result is that the calculation result is usually a decimal value, so that a rounding calculation process is added to perform conversion with high accuracy.

Next, the value Δd ′ is added to the pixel value d (gradation value) to obtain a value d ′ after information embedding.
d ′ = d + Δd ′ (3)
Here, Δd ′ can take both a positive value and a negative value. Therefore, if the value d ′ (= d + Δd ′) after information embedding is smaller than the minimum value dmin of the value that d can take, it is set to dmin. Perform clipping processing. As a result, the value of the pixel falls within a predetermined range even after data embedding.

  By executing the process in the above-described steps, an increase in data capacity due to data embedding can be minimized. In addition, since the number of gradations is not changed, the pattern embedding unit 106 according to the embodiment of the present invention is included in the unit even if the image processing system is integrated into one unit up to the gradation processing. It can be added to the rear stage or installed outside. Therefore, functions can be added flexibly.

[Second Embodiment]
FIG. 5 is a block diagram showing an image processing apparatus according to the second embodiment of the present invention.
In the second embodiment, a low-pass filter (LPF) 108 is newly provided after the gradation processing unit 105 and before the pattern embedding unit 106 in the first embodiment. Is different. Accordingly, the same parts as those in the first embodiment are denoted by the same reference numerals, and detailed description thereof is omitted.

Assume that the input document or image data is preliminarily embedded with specific information by a technique based on this embodiment. In this case, it is necessary to remove old information from the image data to be processed and process only the original image data. The low-pass processing unit 108 removes the old specific information by removing the high band and extracts only the image signal.
Regarding the removal of old specific information, only the image data may be obtained by performing Fourier transform on the signal after gradation processing, extracting the embedded specific information, and removing only the extracted periodic components. . In this case, a processing circuit having a band removal function may be provided instead of the low-pass processing unit 108.

[Third embodiment]
FIG. 6 is a block diagram showing an image processing apparatus according to the third embodiment of the present invention.
In the third embodiment, a smoothing processing unit 109 is newly provided after the gradation processing unit 105 and before the pattern embedding unit 106, and a re-quantization unit 110 is newly provided after the pattern embedding unit 106. This is different from the first embodiment. Accordingly, the same parts as those in the first embodiment are denoted by the same reference numerals, and detailed description thereof is omitted.

The operation of the image processing apparatus shown in FIG. 6 will be described.
The operations from the image data input unit 103 to the gradation processing unit 105 and the operations of the data generation unit 101 and the pattern generation unit 102 are the same as those in the first embodiment. The smoothing processing unit 109 performs a filtering process for the smoothing process on the data after the gradation process. Then, the requantization unit 110 performs the quantization process again on the data in which the pattern is embedded.

The image processing apparatus according to the third embodiment is applied to pattern embedding processing in the case where image data with a small quantization level of gradation processing is targeted.
When the quantization level of the gradation processing is small, the fluctuation range of the pixel value due to the difference of one level is large, so that it is not noticeable even if the amplitude α of the embedded periodic component is increased as compared with the first embodiment. On the other hand, the number n of periodic components that can be embedded is limited. In addition, arrangement | positioning of a periodic component etc. applies to 1st Embodiment.

  FIG. 7 is a diagram illustrating an example of a filtering operator used for the smoothing process. In the figure, the value of the central portion of the 3 × 3 operator indicates the weighting for the target pixel. By setting the weight for the target pixel in the center to a value larger than the weight for the surrounding pixels, the occurrence of extreme blur due to filtering is suppressed.

FIG. 8 is a diagram showing the transition of the image-processed signal. In FIG. 8, the quantization level is set to 2 (= 1, 0) in order to simplify the description.
(1) in FIG. 8 shows a signal after gradation processing by the gradation processing unit 105. (2) in FIG. 8 illustrates a signal after the smoothing process by the smoothing processing unit 109. (3) of FIG. 8 shows a signal after the pattern is embedded by the pattern embedding unit 106 and then re-quantized by the re-quantization unit 110 to the original quantization level.

As shown in FIG. 8, the pattern is not embedded directly in the signal after the gradation processing, but the pattern is embedded after the smoothing processing, thereby suppressing a significant change in the quantization level.
As shown in (3) of FIG. 8, since the pattern is embedded after the smoothing process, the signal is as if the print position is finely corrected by requantization. However, in the case where the data output / display unit 107 controls the number of multi-values by, for example, beam narrowing, which is a well-known method, even in normal processing, printing is performed as in (3) of FIG. The shape is changed as if the position was finely corrected. Therefore, even when configured as in the third embodiment, the original image is rarely damaged.

  According to the method as described above, it is possible to embed information amount data in an image after gradation processing even when the number of gradations used in gradation processing is small and the degree of freedom is low. Become.

[Fourth Embodiment]
FIG. 9 is a block diagram showing an image processing apparatus according to the fourth embodiment of the present invention.
In the fourth embodiment, the image processing apparatus is configured by combining the first embodiment and the third embodiment.

  In the image processing apparatus illustrated in FIG. 9, a switching unit 111 is newly provided, and the functions of the smoothing processing unit 109 and the requantization unit 110 are switched on / off based on data from the gradation processing unit 105. Is different from the third embodiment. Accordingly, the same parts as those in the third embodiment are denoted by the same reference numerals, and detailed description thereof is omitted.

The operation of the image processing apparatus shown in FIG. 9 will be described.
The operations from the image data input unit 103 to the gradation processing unit 105 and the operations of the data generation unit 101 and the pattern generation unit 102 are the same as those in the third embodiment.

The switching unit 111 acquires the number of quantization levels from the gradation processing unit 105.
When the number of quantization levels is equal to or greater than the predetermined value, the switching unit 111 outputs an off signal to the smoothing processing unit 109 and the requantization unit 110. The smoothing processing unit 109 and the requantization unit 110 turn off the respective functions, and output the image signal to the subsequent stage without performing processing.
When the number of quantization levels is smaller than the predetermined value, the switching unit 111 outputs an ON signal to the smoothing processing unit 109 and the requantization unit 110. The smoothing processing unit 109 and the requantization unit 110 turn on the respective functions, execute processing, and output an image signal to the subsequent stage.

According to each embodiment of the present invention described above, there are various effects.
(1) Since it is possible to embed data even after gradation processing is performed, even if a problem such as a change in gradation pattern occurs, it is possible to perform processing without problems by applying the present invention. Become.

(2) Since the processing system is integrated into one unit up to the gradation processing, it is possible to install a data embedding unit as an after-installation and an external attachment even when the change is impossible.

(3) Even if the number of gradations used for gradation processing is small, it is possible to embed data by performing smoothing (high band cut) and requantization.

Note that the present invention is not limited to the above-described embodiment as it is, and can be embodied by modifying the constituent elements without departing from the scope of the invention in the implementation stage.
Further, various inventions can be formed by appropriately combining a plurality of constituent elements disclosed in the embodiment. For example, some components may be deleted from all the components shown in the embodiment. Furthermore, you may combine suitably the component covering different embodiment.

  INDUSTRIAL APPLICABILITY The present invention can be used in an industry for manufacturing an image processing apparatus capable of embedding information in an image that has already been subjected to gradation processing or an image that has been subjected to gradation processing with a small number of gradations.

  DESCRIPTION OF SYMBOLS 1 ... Image processing apparatus, 101 ... Data generation part, 102 ... Pattern generation part, 103 ... Image data input part, 104 ... Image processing part, 105 ... Gradation processing part, 106 ... Embedding part, 107 ... Display part, 108 ... Low-pass processing unit, 109 ... smoothing processing unit, 110 ... re-quantization unit, 111 ... switching unit.

JP-A-7-123244

Claims (12)

Gradation processing means for performing gradation processing on an image;
Signal generating means for generating a data signal representing information different from the image;
Information embedding means for embedding the other information in the image by adding a stripe pattern having a plurality of frequency components corresponding to the data signal to the image signal after the gradation processing. Image processing device.   The image processing apparatus according to claim 1, wherein the information embedding unit does not change the number of gradations of the image before and after embedding the other information.   The information embedding means adjusts the fluctuation value to an integer multiple of one level used for gradation processing when the pixel value of the image changes by embedding the other information. The image processing apparatus according to claim 2.   The image processing apparatus according to claim 2, further comprising a high frequency component removing unit that is provided in a preceding stage of the information embedding unit and removes a high frequency component of the image signal. A smoothing unit that is provided in a preceding stage of the information embedding unit and executes a smoothing process on the image signal;
Re-quantization means provided at a subsequent stage of the information embedding means for quantizing the image signal in which the other information is embedded so as to have the same multi-value number as the multi-value number of the gradation processing The image processing apparatus according to claim 1, further comprising:   The smoothing unit and the re-quantization unit are configured to execute a smoothing operation and a re-quantization operation, respectively, when the multi-value number of the gradation processing is equal to or greater than a predetermined value. 6. The image processing apparatus according to claim 5, wherein when the number of values is smaller than a predetermined value, the input image signal is output as it is without performing each of the operations. Perform gradation processing on the image,
Generate a data signal that represents information different from the image,
An image processing method comprising embedding the other information in the image by adding a striped pattern having a plurality of frequency components corresponding to the data signal to the image signal after the gradation processing.   8. The image processing method according to claim 7, wherein in embedding the other information, the number of gradations of the image is not changed before and after embedding of the other information.   9. The value of the variation is adjusted to an integer multiple of one level used for gradation processing when the pixel value of the image changes due to the embedding of the other information. Image processing method.   9. The image processing method according to claim 8, wherein a high-frequency component of the image signal is removed before embedding the other information. Before performing the embedding of the other information, perform a smoothing process on the image signal,
After performing the embedding of the other information, the image signal in which the other information is embedded is quantized so that the multi-value number is the same as the multi-value number of the gradation processing. The image processing method according to claim 7.   The smoothing operation and the re-quantization operation are executed when the multi-value number of the gradation processing is greater than or equal to a predetermined value, and each operation is performed when the multi-value number of the gradation processing is smaller than the predetermined value. The image processing method according to claim 11, wherein the image signal input without being output is output as it is.

Images