KR101007291B1 - Method for hiding geographical information in image data, and computer readable recording medium storing geographical information hiding program - Google Patents

Abstract

The present invention relates to a method for concealing geographic information in an image data and a computer-readable recording medium storing a geographic information concealment program. The method for concealing geographic information in an image data according to the present invention comprises: receiving GPS data; step; Extracting latitude and longitude values from the GPS data; Dividing the target image into which the watermark is to be inserted in units of 8 × 8 pixels, and setting a plurality of image blocks; Generating a watermark based on the number of the plurality of image blocks, the latitude value, and the longitude value; And in the process of encoding the target image data in JPEG (Joint Photographic Experts Group), by adjusting the size of the first or second color difference component of each of the plurality of image blocks according to the bit value of the watermark, Generating the concealed image data. Bits of the watermark correspond to the plurality of image blocks, respectively. According to the present invention, the invisibility and robustness of the image in which the geographic information is hidden can be increased, and the security of the location information can be guaranteed.

Geographic information, watermarks, luminance components, chrominance components

Description

Method for hiding geographical information in image data, and computer readable recording medium storing geographical information hiding program}

The present invention relates to a method for concealing geographic information, and more particularly, to a method of concealing geographic information in image data and a computer-readable recording medium storing a geographic information concealment program.

Recently, the emergence of new concepts and new convergence technologies represented by web 2.0 Software as a Service (SaaS), digital conversion (digital convertgence), and the like, suggests a new paradigm of IT (Information Technology) environment. In addition, the emergence of mesh-up, a technology and service that provides a new form of user-oriented services and applications by combining mutually independent technologies and contents with little correlation, has been a one-way web 1.0 SaaS. Is providing a significantly different user-participatory IT environment. To keep pace with this trend, the location and geographic information of the Global Positioning System (GPS) is no longer limited to digital information for cartography, but rather embeds location coordinates into all multimedia content, such as photos and videos that represent regions, It is evolving as information for proving this. FIG. 1 shows a general satellite observation picture in which pictures to which location information is applied are provided to provide a mashup service. As shown in FIG. 2, the picture to which the location information of FIG. 1 is applied, the GPS reception and image acquisition device 10, converts the acquired image 30 into an image file in a specific format (for example, JPG or TIF). The location information received from the GPS satellite 20 can be obtained by inserting the location information received from the GPS satellite 20 into an image file header of an EXIF (Exchangeable image file format) by a geotagging process. Here, the GPS reception and image acquisition device 10 may include, for example, a mobile phone or a camera equipped with a GPS receiver.

However, since inserting geographic information into multimedia contents is a technology that has just been introduced, there are problems in terms of user convenience and information protection. From a user convenience point of view, the existing method of inserting geographic information requires manual manual work by the user. In addition, from the information protection point of view, since the location information tagged in the image file header is important information of the individual, for example, the traces of the individual are exposed and the privacy of the individual is likely to be violated. In addition, since the location information tagged in the image file header may be easily damaged or maliciously modified by others, there is a great risk of, for example, alibi manipulation at a crime scene. On the other hand, when the format of the image file tagged with the location information is converted, there is a fear that the location information is lost or becomes impossible to recover.

Accordingly, a technical problem to be achieved by the present invention is a bit value of a watermark generated based on latitude and longitude values according to GPS location information in the process of encoding target image data in JPEG (Joint Photographic Experts Group) encoding. According to the present invention, a method of concealing geographic information in image data in which invisibility and robustness is increased and security of location information can be ensured by adjusting the size of the first or second color difference component included in the target image data. To provide.

Another technical problem to be solved by the present invention is to provide a method for encoding the target image data according to a bit value of a watermark generated based on latitude and longitude values according to GPS location information. By adjusting the size of the first or second color difference component, an invisibility and robustness can be increased, and a computer-readable recording medium storing a geographic information concealment program capable of ensuring the security of location information can be provided.

[수학식 2]

According to an aspect of the present invention, there is provided a method of concealing geographic information in an image data, the method comprising: receiving GPS data; Extracting latitude and longitude values from the GPS data; Dividing a target image into which a watermark is to be inserted in units of 8x8 pixels, and setting a plurality of image blocks; Generating a watermark based on the number of the plurality of image blocks, the latitude value, and the longitude value; And in the process of encoding the target image data in JPEG (Joint Photographic Experts Group), by adjusting the size of the first or second color difference component of each of the plurality of image blocks according to the bit value of the watermark, Generating hidden image data, wherein bits of the watermark correspond to the plurality of image blocks, and generating the watermark comprises the latitude value and the longitude value respectively formed of a set number of digits. Calculating a repetition number of each digit of the latitude value and the longitude value based on Equation 1 and Equation 2 below in order to set the importance of each digit differently; Generating a bit string by converting the latitude value and the longitude value of which the number of digits are increased by the number of repetitions into binary-coded decimal numbers (BCD); And rearranging the bits of the bit string based on a preset key value defining a sequence of bits to generate a watermark.
[Equation 1]

[Equation 2]

[수학식 2]

According to another aspect of the present invention, there is provided a computer-readable recording medium storing a geographical information concealment program, the method comprising: receiving GPS data; Extracting latitude and longitude values from the GPS data; Dividing the target image into which the watermark is to be inserted in units of 8x8 pixels, and setting a plurality of image blocks; Generating a watermark based on the number of the plurality of image blocks, the latitude value, and the longitude value; And in the process of JPEG encoding the target image data, adjusting the size of the first or second color difference component of each of the plurality of image blocks according to the bit value of the watermark, thereby generating image data concealing geographic information. A computer-readable recording medium storing a geographic information concealment program for executing a step, wherein bits of the watermark correspond to the plurality of image blocks, and generating the watermark comprises a predetermined number of digits. In order to set the importance of each latitude of each of the latitude value and the longitude value differently, the number of repetitions of each latitude of the latitude value and the longitude value is based on Equation 1 and Equation 2 below. Calculating; Generating a bit string by converting the latitude value and the longitude value of which the number of digits are increased by the number of repetitions into binary-coded decimal numbers (BCD); And rearranging the bits of the bit string based on a preset key value defining a sequence of bits to generate a watermark.
[Equation 1]

[Equation 2]

As described above, the computer-readable recording medium storing the geographic information in the image data and the geographic information concealment program according to the present invention, GPS in the process of encoding the target image data in the JPEG (Joint Photographic Experts Group), GPS Since the size of the first or second color difference component included in the target image data is adjusted according to the bit value of the watermark generated based on the latitude value and the longitude value according to the position information of, the geographic information is hidden. The invisibility and robustness of the rendered image can be increased. In addition, since the geographic information is concealed in the image in the form of a watermark, it is impossible to alter or modify the geographic information by another person, thereby ensuring the security of the location information.

Hereinafter, with reference to the accompanying drawings will be described a preferred embodiment of the present invention. However, the present invention is not limited to the embodiments disclosed below, but may be embodied in various different forms, and only the embodiments are intended to complete the disclosure of the present invention and to those skilled in the art. It is provided for complete information.

3 is a flowchart illustrating a process of hiding geographic information in image data according to an embodiment of the present invention. The GPS (Global Position System) Receiving and Image Acquisition Device 100 (see FIG. 8) receives the GPS data GDATA (step 1100). The GPS reception and image acquisition apparatus 100 may be, for example, a mobile phone, a digital camera, a camcorder, a digital video recorder (DVR), a scanner, a personal digital assistant (PDA), a portable media player (PMP) equipped with a GPS receiver. It can be implemented by any one of the computer. The GPS reception and image acquisition apparatus 100 receives GPS data GDATA, for example, in ASCII (ASCII) code format of 256 bytes per second.

The GPS reception and image acquisition apparatus 100 extracts a latitude value LAT and a longitude value LON from the GPS data GDATA (step 1200). At this time, the GPS reception and image acquisition apparatus 100 extracts latitude value LAT and longitude value LON using Global Position Recommended Minimum specific GPS / Transit data (GPRMC) data included in the GPS data GDATA. do. GPRMC data includes information such as time, latitude, longitude, height, and altitude.

The GPS reception and image acquisition apparatus 100 may display the target image P (see FIG. 11) having the size of A × B (for example, 512 × 512) pixels into which the watermark WM1 will be inserted (see FIG. 11). A plurality of image blocks P1 to PS (S is J × K, J and K are integers) (see FIG. 13A) are divided in units of (step 1300). Thereafter, the GPS receiving and image obtaining apparatus 100 generates a watermark based on the number S of the plurality of image blocks, the latitude value LAT, and the longitude value LON (step 1400). . The bits of the watermark WM1 correspond to the plurality of image blocks P1 to PS, respectively. That is, the number of bits of the watermark WM1 is equal to the number of image blocks P1 to PS.

Afterwards, the GPS reception and image acquisition apparatus 100 encodes the target image data PDATA in the process of encoding the Joint Photographic Experts Group (JPEG), and the color difference component Cb of each of the plurality of image blocks P1 to PS is generated. By adjusting the size of (Chrominance Blue) or Cr (Chrominance Red) (first or second color difference component) according to the bit value of the watermark WM1, the image data P 'concealed in geographic information (Fig. 8) (step 1500).

4 is a detailed flowchart of the generation step 1400 shown in FIG.

The GPS reception and image acquisition apparatus 100 sets latitude values LAT and longitude values LON in order to set different degrees of importance for each digit of each of the latitude value LAT and the longitude value LON respectively. The number of repetitions for each digit of is calculated (step 1401). The number of repetitions for each digit of the latitude value LAT and the longitude value LON may be set to an equal order structure represented by Equation 1 below.

In Equation 1, the first value a is a number corresponding to the number of repetitive insertions of the least significant digit of the latitude value or the longitude value, and varies according to the number S of the entire image blocks. The first value (a) is the minimum value that is the criterion of the watermark WM1 when the watermark WM1 is extracted from the image into which the watermark WM1 is inserted.

Meanwhile, the first term value a and the tolerance value d may be calculated by Equation 2 below.

The left side value S / 8 of Equation 2 indicates the total number of insertions of the latitude value or the longitude value. In the left-side value (S / 8), '8' represents 2 (represents two geographies of latitude and longitude) × 4 (latitude or longitude values in decimal are converted by binary-coded decimal numbers (BCD) conversion. When converted to binary, the number of bits in each digit).

The first value (a) corresponds to the minimum odd integer value corresponding to approximately 2% of the total number of insertions (S / 8). For example, when the number S of image blocks is' 4096 ', the total number of inserts S / 8 is 512, and 2% of the total number of inserts S / 8 is '10 .24'. Since the first term value a is an integer value of the minimum odd number, it is determined as' 11 'which is the minimum odd value larger than '10 .24'.

In addition, the tolerance value d may be determined as an integer value of the minimum odd number satisfying Equation (2). For example, when the first value (a) is' 11 ', the number of image blocks (S) is 4096, and the latitude value (LAT) or longitude value (LON) has 9 digits, '11 .472222≥d' Is calculated. As a result, the tolerance value d is determined to be '11' because it is a minimum odd integer value satisfying '11 .472222? D '.

For example, when the latitude value LAT is 03508.0658 and the longitude value LON is 12906.1795, each digit number of the latitude value LAT and the longitude value LON may be represented as shown in the table below.

Place number 9 8 7 6 5 4 3 2 One Latitude value 0 3 5 0 8 0 6 5 8 Longitude value One 2 9 0 6 One 7 9 5

In addition, when the first value (a) is '11', the tolerance value (d) is '11', and the number (S) of image blocks is '4096', [n] in [Equation 1] When the digits of 1] are substituted and calculated, the calculation result of the number of repetitions of each digit of the latitude value (LAT) and the longitude value (LON) is as shown in [Table 2].

Repeat count 99 88 77 66 55 44 33 22 11 Latitude value 0 3 5 0 8 0 6 5 8 Longitude value One 2 9 0 6 One 7 9 5

When the number of repetitions shown in [Table 2] is applied to the latitude value (LAT) and the longitude value (LON) respectively, the latitude value (LAT) is changed from 9 to 495 digits, and the longitude value (LON) is also 9 digits. The number changes to 495 digits. As a result, the sum of the total digits of the latitude value LAT and the longitude value LON is '990'.

Meanwhile, since the total number of insertions (S / 8) for the latitude value (LAT) is '512' and the total number of insertions (S / 8) for the longitude value (LON) is '512', the latitude value (LAT) The number of insertions for the entire longitude value LON is '1024'.

Here, since the total number of insertions '1024'-the total number of digits '990' = 34, the specific positions of each of the latitude value LAT and the longitude value LON must be repeatedly inserted 17 times. For the robustness of the watermark WM1, it is preferable that the most significant digit of each of the latitude value LAT and the longitude value LON is repeatedly inserted 17 times. As described above, when the most significant digit of each of the latitude value LAT and the longitude value LON is inserted 17 times more repeatedly, [Table 2] may be represented as shown in [Table 3] below.

Repeat count 116 88 77 66 55 44 33 22 11 Latitude value 0 3 5 0 8 0 6 5 8 Longitude value One 2 9 0 6 One 7 9 5

The reason why the number of repetitions of the upper digit is set larger than the number of repetitions of the lower digit is as follows. This means that the watermark should not be removed or destroyed.) To ensure the reliability of the location information even if the image data concealed by the geographic information is subjected to various attacks later.

For example, the latitude value of geographic information inserted as a watermark in specific image data is 35 degrees 08 minutes 39.48 seconds, the longitude value is 129 degrees 6 minutes 10.77 seconds, and the geographic information detected from the attacked image data (i.e., , Latitude value of 35 degrees, 08 minutes, 39.48 seconds, and longitude value of 128 degrees, 06 minutes, 10.77 seconds. In this case, there is a difference of 1 degree between the inserted hardness value and the detected hardness value, but 1 degree corresponds to approximately 90 km in terms of the actual distance value, so that the position of the detected hardness value is completely different from the position of the inserted hardness value. Can be. Therefore, among the digits of the latitude value LAT and the longitude value LON, when an error occurs, the number of repetitions of the upper digit which greatly affects the reliability of the location information is increased, thereby increasing the error of the upper digit value. It is necessary to reduce the occurrence of.

Thereafter, the GPS reception and image acquisition apparatus 100 converts the latitude value (LAT) (512 digits) and the longitude value (LON) (512 digits) of which the number of digits have been increased by the number of repetitions to binary-coded decimal numbers (BCD). Then, a bit string (4096 total positions) is generated (step 1402). 9 illustrates an example of a bit string of latitude values LAT to which the number of repetitions of [Table 3] is applied.

The GPS reception and image acquisition apparatus 100 rearranges the bits of the bit string generated in step 1042 based on a preset key value BK that defines the order of the bits, thereby generating a watermark WM1. (Step 1403). The key value BK relates to the order of the bits to cause the order of the bits (2048) of the latitude value (LAT) and the bits (2048) of the longitude value (LON) to be randomly shuffled together. Represent information.

Referring to FIG. 10, this will be described in more detail. For simplicity of explanation, the process by which 16 bits are rearranged is described. For example, when the key value BK is "5 13 7 0 4 9 15 6 10 3 1 14 8 11 12 2", and the bit string is "0011 0101 0000 1000", the key value BK is rewritten according to the key value BK. The arranged bit strings are " 1010 0000 0100 0011 ".

5 is a detailed flowchart of the generation step 1500 shown in FIG.

The GPS reception and image acquisition apparatus 100 selects one of the plurality of image blocks P1 to PS (see FIG. 13A) (step 1501). The GPS reception and image acquisition apparatus 100 extracts luminance (Y) component values and color difference (Cb, Cr) component values (first and second color difference component values) of the selected image block (for example, P1) ( Step 1502) (see Figure 13B).

Although not shown in FIG. 5, the information of the user who generated the target image data P as an additional watermark WM2 may be inserted in the luminance Y component values extracted in step 1502. The information of the user may include, for example, at least one of a serial number of a device generating the target image data P, date and time information of generating the target image data P, and a phone number of the user.

The GPS reception and image acquisition apparatus 100 may convert luminance (Y) component values into luminance (Y) frequency coefficient values (YF _{1,0,0 to} YF _1,7 ) by a discrete cosine transform (DCT). ₇ ), color difference (Cb) component values (first color difference component values) to color difference (Cb) frequency coefficient values (first color difference frequency coefficient values) (CbF _{1,0,0 to} CbF _1,7,7 ), Color difference (Cr) component values (second color difference component values) are converted into color difference (Cr) frequency coefficient values (second color difference frequency coefficient values) (CrF _{1,0,0 to} CrF _1,7,7 ), respectively (step 1503).

The GPS reception and image acquisition apparatus 100 may include luminance frequency coefficient values YF _{1,0,0 to} YF _1,7,7 , color difference Cb frequency coefficient values CbF _{1,0,0 to} CbF _{1,7, 7} ) and chrominance (Cr) frequency coefficient values (CrF _{1,0,0 to} CrF _1,7,7 ) are rearranged by a zigzag scan (step 1504). 15 is a conceptual diagram illustrating a process in which chrominance Cb frequency coefficient values CbF _{1,0,0 to} CbF _1,7,7 are zigzag scanned, and in FIG. 17, color difference (Cr) frequency coefficient values ( CrF _1,0,0 ~CrF _1,7,7) is a conceptual diagram illustrating a process of zigzag scanning is shown. In Figures 15 and 17, for the sake of simplicity, the respective frequency coefficient values are indicated by dots only.

The reason why the zigzag scan process is performed for each frequency coefficient value is that the zigzag scan process efficiently sets each frequency coefficient value to a low frequency component (i.e., a direct current component) and a high frequency component (i.e., an alternating current component). By separately arranging), the quantization process to be performed later can be easily performed.

FIG. 18 illustrates an arrangement state after color difference Cb frequency coefficient values CbF _{1,0,0 to} CbF _1,7,7 are zigzag scanned, and FIG. 19 illustrates color difference Cr frequency coefficient values CrF _1, The arrangement state after _{0,0 to} CrF _1,7,7 is zigzag scanned is shown. On the other hand, "1" in "1,0,0" of the color difference (Cb) frequency coefficient value (CbF _1,0,0 ) represents the number of the image block (that is, P1), and "0,0" represents the color difference ( Cb) represents a row and column number in which frequency coefficient values CbF _{1,0,0 to} CbF _1,7,7 are arranged. The color difference (Cb) frequency coefficient values of the 0th row and 0th column of the S-th image block may be expressed as "CbF _{S, 0,0} ". Similarly, "1" in "1,0,0" of the color difference (Cr) frequency coefficient value (CrF _1,0,0 ) represents the number of the image block (ie, P1), and "0,0" Color difference (Cr) represents a row and column number in which frequency coefficient values CrF _{1,0,0 to} CrF _1,7,7 are arranged.

Thereafter, the GPS reception and image acquisition apparatus 100 rearranges the rearranged luminance frequency coefficient values (YF _{1,0,0 to} YF _1,7,7 ) and the color difference (Cb) frequency coefficient values (CbF _1,0,0). ˜CbF _1,7,7 and color difference (Cr) frequency coefficient values CrF _{1,0,0 to} CrF _1,7,7 are respectively quantized (step 1505). In step 1505, quantization is performed using a preset quantization table (see FIG. 20).

The quantized luminance frequency coefficient values YF _1,0,0 'to YF _1,7,7 ' are obtained by _{converting the} luminance frequency coefficient values YF _{1,0,0 to} YF _1,7,7 to the values QF of the quantization table. _0,0, ˜QF _7,7 ) respectively. The quantized color difference (Cb) frequency coefficient values (CbF _1,0,0 ′ to CbF _1,7,7 ′) represent the color difference (Cb) frequency coefficient values (CbF _{1,0,0 to} CbF _1,7,7 ). Correspond to the values divided by the values of the quantization table (QF _0,0, ˜QF _7,7 ), respectively. For example, the quantized color difference Cb frequency coefficient value CbF _1,0,0 ′ may be expressed as Equation 3 below.

The quantized color difference (Cr) frequency coefficient values (CrF _1,0,0 'to CrF _1,7,7 ') are used to determine the color difference (Cr) frequency coefficient values (CrF _{1,0,0 to} CrF _1,7,7 ). Corresponding to the values divided by the values of the quantization table (QF _{0,0 to} QF _7,7 ), respectively. For example, the quantized color difference (Cr) frequency coefficient value CrF _1,0,0 'may be expressed as shown in Equation 4 below.

The GPS reception and image acquisition apparatus 100 may include a DC component value (ie, CbF _1,0,0 ′) of quantized color difference (Cb) frequency coefficient values CbF _{1,0,0 to} CbF _1,7,7 . the quantized color difference (Cr) frequency coefficient values of the DC component values (CrF _1,0,0 ~CrF _1,7,7) (that is, CrF _1,0,0 ') between the difference value (QCDV; chrominance quantized difference value) Is calculated (step 1506). The difference value QCDV may be expressed as Equation 5 below.

For example, the values of the quantization table QF _0,0 are 3, and the DC component values CbF _{1, of the} quantized color difference Cb frequency coefficient values CbF _{1,0,0 to} CbF _1,7,7 when _{0, 0)} is 45, and the quantized color difference (Cr) frequency coefficient values (CrF _1,0,0 ~CrF DC component value _1,7,7) (CrF _1,7,7) 36 days, primary The value QCDV is calculated as three.

The GPS reception and image acquisition apparatus 100 may quantize the color difference Cb frequency such that the difference value QCDV satisfies a condition set according to the corresponding bit value of the watermark WM1 corresponding to the selected image block P1. DC component values (ie, CbF _1,0,0 ') of the coefficient values (CbF _1,0,0 ' to CbF _1,7,7 ') or quantized color difference (Cr) frequency coefficient values (CrF _1,0, regulates ₀ '~CrF _1,7,7') DC component values (i.e., CrF _1,0,0 'a) (step 1507).

Thereafter, the GPS reception and image acquisition apparatus 100 determines whether the image block selected in step 1501 is the last image block (step 1508). If the image block selected in step 1501 is not the last image block, the GPS reception and image acquisition apparatus 100 may determine the DC component value and the quantized color difference Cr of the quantized color difference Cb frequency coefficient values of each of the entire image blocks. Until the DC component values of the frequency coefficient values are adjusted, the GPS reception and image acquisition apparatus 100 repeatedly performs the operations of steps 1501 to 1508.

In addition, when the image block selected in step 1501 is the last image block, the GPS reception and image acquisition apparatus 100 may adjust the color difference (Cb) frequency coefficient values of the quantized luminance frequency coefficient values and the DC component values of all the image blocks. And entropy code the quantized color difference (Cr) frequency coefficient values (step 1509). As a result, image data P '(see FIG. 8) concealing geographic information (i.e. latitude and longitude values) is generated.

FIG. 6 is an example of a detailed flowchart of the adjustment step 1507 shown in FIG.

The GPS reception and image acquisition apparatus 100 sets a divisor value M of a difference value QCDV between DC component values of the quantized color difference (Cb) frequency coefficient values and the quantized color difference (Cr) frequency coefficient values. (M = 4H, H is an integer greater than 0 for determining the embedding strength of the watermark), and the remaining value is obtained (step 1511). Here, the operation for obtaining the remaining value T may be performed by the mod operation shown in Equation 6 below. The divisor value M may be used later as a key value for extracting a watermark from the image data PDATA 'where the geographic information is hidden.

For example, when the difference value QCDV is 3 and the divisor value M is 8, by Equation 6, the remaining value T is calculated to be 3.

The GPS reception and image acquisition apparatus 100 determines whether the corresponding bit value of the watermark WM1 corresponding to the selected image block (ie, to be inserted) is 1 (step 1512). The bits WSW (S-1) ... W2W1 of the watermark WM1 correspond to the image blocks P1 to PS, respectively, as shown in Table 3 below.

Bits of Watermark WM1 WS W (S-1) ... W2 W1 Image blocks P1 P2 ... P (S-1) PS

When the corresponding bit value (e.g., WS) of the watermark corresponding to (e.g., inserted into) the selected image block (e.g., P1) is 1, the GPS reception and image acquisition apparatus 100 is M / 2. It is determined whether or not <rest value T < M (or -M / 2 < rest value T < 0) (step 1513).

When M / 2 <rest value T < M (or -M / 2 <rest value T < 0), the GPS reception and image acquisition device 100 is a dashed arrow A1 (or A2) of FIG. Quantized color difference Cb frequency coefficient values (first color difference) such that 0 <remaining value T <M / 2 (or -M <remaining value T <-M / 2) is satisfied. Decrease the DC component value of the frequency coefficient values) or increase the DC component value of the quantized color difference (Cr) frequency coefficient values (second chrominance frequency coefficient values) (step 1514). For example, when the difference value QCDV is 4 and the divisor M is 8, the GPS reception and image acquisition apparatus 100 may satisfy the quantized color difference Cb so as to satisfy 0 <rest value T <4. ) Decrease the DC component value of the frequency coefficient values (first chrominance frequency coefficient values) by -1 (or -2) or reduce the DC component value of the quantized color difference (Cr) frequency coefficient values (second chrominance frequency coefficient values). Increase by +1 (or +2).

Further, when M / 2 <remaining value T <M (or -M / 2 <remaining value T <0), that is, 0 <remaining value T <M / 2 (or -M < When the remaining values (T) <-M / 2), the GPS reception and image acquisition apparatus 100 may determine the DC component values and the quantized color differences () of the quantized color difference Cb frequency coefficient values (the first color difference frequency coefficient values). Cr) The DC component values of the frequency coefficient values (second chrominance frequency coefficient values) are kept as is (step 1515).

On the other hand, in step 1512, when the corresponding bit value (e.g., WS) of the watermark corresponding to (e.g., inserted) the selected image block (e.g., P1) is not 1, i.e., 0, The GPS reception and image acquisition apparatus 100 determines whether 0 <rest value T <M / 2 (or -M <rest value T <-M / 2) or not (step 1516).

When 0 <rest value T <M / 2 (or -M <rest value T <-M / 2), the GPS reception and image acquisition device 100 is a solid arrow A3 (or A4) in FIG. Quantized color difference (Cb) frequency coefficient values (first color difference) to satisfy M / 2 <remaining value (T) <M (or -M / 2 <remaining value (T) <0), as indicated by Increase the DC component value of the frequency coefficient values) or decrease the DC component value of the quantized color difference (Cr) frequency coefficient values (second chrominance frequency coefficient values) (step 1517). For example, when the difference value QCDV is 3 and the divisor M is 8, the GPS reception and image acquisition apparatus 100 may have 4 <remaining value T <8 (or −4 <remaining value T Increase the DC component value of the quantized color difference (Cb) frequency coefficient values (first color difference frequency coefficient values) by +2 or satisfy the quantized color difference (Cr) frequency coefficient values (second color difference frequency) to satisfy) <0). Reduce the DC component value of the coefficient values) by -2. At this time, the DC component values of the quantized color difference (Cb) frequency coefficient values (first color difference frequency coefficient values) are increased by +1, and the DC component values of the quantized color difference (Cr) frequency coefficient values (second color difference frequency coefficient values) are increased. May be reduced to -1.

In addition, when 0 <rest value T <M / 2 (or -M <rest value T <-M / 2), that is, M / 2 <rest value T <M (or -M / When 2 <the rest value T <0), the GPS reception and image acquisition apparatus 100 determines the DC component value and the quantized color difference Cr of the quantized color difference Cb frequency coefficient values (first color difference frequency coefficient values). ) Maintains the DC component values of the frequency coefficient values (second chrominance frequency coefficient values) (step 1518).

FIG. 7 is another example of a detailed flowchart of the adjusting step 1507 shown in FIG.

The GPS reception and image acquisition apparatus 100 sets the difference value QCDV between the DC component values of the quantized color difference (Cb) frequency coefficient values and the quantized color difference (Cr) frequency coefficient values. 4H, where H is an integer greater than 0 that determines the embedding strength of the watermark), and the remaining value is obtained (step 1521). Here, the operation for obtaining the remaining value T can be executed by the mod operation shown in Equation 6 described above. The GPS reception and image acquisition apparatus 100 determines whether or not the corresponding bit value of the watermark WM1 corresponding to the selected image block (ie, to be inserted) is 1 (step 1522).

When the corresponding bit value (e.g., WS) of the watermark corresponding to (e.g., inserted into) the selected image block (e.g., P1) is 1, the GPS reception and image acquisition apparatus 100 is M / 2. It is determined whether < remaining value T < M (or -M / 2 < remaining value T < 0) (step 1523).

When M / 2 <rest value T < M (or -M / 2 <rest value T < 0), the GPS reception and image acquisition device 100 is a dashed arrow A1 (or A2) of FIG. Quantized color difference Cb frequency coefficient values (first color difference) such that 0 <remaining value T <M / 2 (or -M <remaining value T <-M / 2) is satisfied. Decrease the DC component value of the frequency coefficient values) or increase the DC component value of the quantized color difference (Cr) frequency coefficient values (second chrominance frequency coefficient values) (step 1524).

Further, when M / 2 <remaining value T <M (or -M / 2 <remaining value T <0), that is, 0 <remaining value T <M / 2 (or -M < When the remaining value (T) <-M / 2), or after step 1524 is executed, the GPS reception and image acquisition device 100 determines whether the remaining value T = M / 4 (or -3M / 4). It is determined whether or not (step 1525).

When the remaining value T = M / 4 (or −3M / 4), the GPS reception and image acquisition apparatus 100 may determine the DC component values of the quantized color difference Cb frequency coefficient values (first color difference frequency coefficient values). And retains the DC component values of the quantized color difference (Cr) frequency coefficient values (second color difference frequency coefficient values) (step 1526).

When the remaining value T is not M / 4 (or -3M / 4), the GPS reception and image acquisition apparatus 100 quantizes such that the remaining value T is M / 4 (or -3M / 4). Increase or decrease the DC component value of the chrominance Cb frequency coefficient values (first chrominance frequency coefficient values) or the quantized chrominance Cr frequency coefficient values (second chrominance frequency coefficient values) (step 1527). The reason for adjusting so that the remaining value (T) = M / 4 (or -3M / 4) is that the remaining value (T) is 0 <remaining value (T) <M / 2 (or -M <remaining value (T)). By moving to the center of the range of <-M / 2), the toughness is further increased.

On the other hand, in step 1522, when the corresponding bit value (e.g., WS) of the watermark corresponding to (e.g., inserted) the selected image block (e.g., P1) is not 1, i.e., 0, The GPS reception and image acquisition apparatus 100 determines whether 0 <rest value T <M / 2 (or -M <rest value T <-M / 2) or not (step 1528).

When 0 <rest value T <M / 2 (or -M <rest value T <-M / 2), the GPS reception and image acquisition device 100 is a solid arrow A3 (or A4) in FIG. Quantized color difference (Cb) frequency coefficient values (first color difference) to satisfy M / 2 <remaining value (T) <M (or -M / 2 <remaining value (T) <0), as indicated by Increase the DC component value of the frequency coefficient values) or decrease the DC component value of the quantized color difference (Cr) frequency coefficient values (second chrominance frequency coefficient values) (step 1529).

In addition, when 0 <rest value T <M / 2 (or -M <rest value T <-M / 2), that is, M / 2 <rest value T <M (or -M / When 2 <rest value T <0), or after step 1529 is executed, whether the GPS reception and image acquisition apparatus 100 has the remaining value T = 3M / 4 (or -M / 4) Determine (step 1530).

When the remaining value T = 3M / 4 (or −M / 4), the GPS reception and image acquisition apparatus 100 may determine the DC component values of the quantized color difference Cb frequency coefficient values (first color difference frequency coefficient values). And maintains the DC component values of the quantized color difference Cr frequency coefficient values (second color difference frequency coefficient values) (step 1531).

When the remaining value T is not 3M / 4 (or -M / 4), the GPS reception and image acquisition apparatus 100 quantizes such that the remaining value T is 3M / 4 (or -M / 4). Increase or decrease the DC component value of the color difference (Cb) frequency coefficient values (first color difference frequency coefficient values) or the quantized color difference (Cr) frequency coefficient values (second color difference frequency coefficient values) (step 1532). The reason for adjusting so that the remaining value (T) = 3M / 4 (or -M / 4) is that the remaining value (T) is M / 2 <remaining value (T) <M (or -M / 2 <remaining value ( This is to increase the toughness by moving to the center of the range of T) <0).

22 is a diagram illustrating JPEG encoded images and images in which geographic information is concealed in a JPEG encoding process according to a geographic information concealment process according to the present invention. As an objective measure of invisibility of the watermark, PSNR (Peak Signal to Noise Ratio) is used. (A) and (c) of FIG. 22 are JPEG encoded images without a watermark inserted (i.e., no geographic information is concealed), and (b) and (d) are inserted with a watermark (i.e., Geographic information is concealed) to represent a JPEG encoded image.

As referred to in the drawing, the images (b) and (d) with the watermark embedded therein are deteriorated in image quality without a perceptual difference when compared with the images (a) and (c) without the watermark embedded. It is confirmed that does not occur. In Table 5, PSNRs of images (a) and (c) without a watermark and images (b) and (d) with a watermark are displayed.

division No watermark inserted Insert watermark No watermark inserted Insert watermark image (a) image (b) image (c) image (d) image PSNR [dB] 27.34 27.84 29.68 31.10

As referenced in Table 5, the PSNR values of JPEG encoded images (a, c) with watermarks inserted are close to the PSNR values of JPEG encoded images (b, d) without watermarks inserted. It can be seen.

FIG. 23 is a diagram illustrating images according to a result of executing various attack experiments on the image shown in FIG. 22B. FIG. 23A shows an image when JPEG encoding is performed once more on the image of FIG. 22B (that is, when a JPEG attack is applied). FIG. 23B shows an image when an attack by median filtering is applied to the image of FIG. 22B. The median filtered image is transformed into a softer image than the original image.

FIG. 23C shows an image when an attack caused by sharpening filtering is applied to the image of FIG. 22B. The sharpened filtered image is clearer and clearer than the original image. In the case of sharpening filtering, the smaller the alpha value (see FIG. 24), the sharper the image. FIG. 23D shows an image when a reduction attack is applied to the image of FIG. 22B.

FIG. 24 is a table showing result values related to the attack experiment shown in FIG. 23.

As referenced in the table of FIG. 24, bit errors occurred in all attacks, but latitude and longitude values could be detected in each attacked image. When the JPEG compression ratio (i.e. Quality) was 40, a difference of 0.0110 occurred between the inserted hardness value and the detected hardness value, but an error occurred at the lower digit of the hardness value, and this difference is converted to the actual distance. It can be seen that the relatively small error of 17m.

FIG. 25 is a table showing values of execution results of various attack experiments on the image shown in FIG. 22D. The table of FIG. 25 shows the case where JPEG encoding is performed once more (ie, when a JPEG attack is applied) to the image of FIG. 22D, when an attack by median filtering process is applied, and sharpening. The result values when an attack by a sharpening filtering process is applied and when a reduction (or enlargement) attack is applied are shown. Bit errors occurred in all attacks, but latitude and longitude values could be detected in each attacked image. In addition, when the JPEG compression ratio (i.e. Quality) is 40 to 80 and the median mask is 7x7 and 9x9, there is a slight difference between the inserted latitude value or longitude value and the detected latitude value or longitude value. Although the error occurred in the lower digit of the latitude value or the longitude value, it can be seen that a relatively short distance error occurs.

The above embodiments are for explaining the present invention, and the present invention is not limited to these embodiments, and various embodiments are possible within the scope of the present invention. In addition, although not described, equivalent means will also be referred to as being incorporated in the present invention. Therefore, the true scope of the present invention will be defined by the claims below.

1 is a view illustrating a general satellite observation picture displaying pictures to which location information is applied.

FIG. 2 is a conceptual diagram schematically illustrating an example of a process of generating the photos illustrated in FIG. 1.

3 is a flowchart illustrating a process of hiding geographic information in image data according to an embodiment of the present invention.

4 is a detailed flowchart of the watermark generation step 1400 illustrated in FIG. 3.

5 is a detailed flowchart of an image data generation step 1500 in which geographic information shown in FIG. 3 is concealed.

FIG. 6 is an example of a detailed flowchart of the adjustment step 1507 shown in FIG.

FIG. 7 is another example of a detailed flowchart of the adjusting step 1507 shown in FIG.

FIG. 8 is a conceptual diagram illustrating a process of hiding geographic information in the image data shown in FIG. 3.

FIG. 9 is a conceptual diagram for describing a bit string generation step 1402 illustrated in FIG. 4.

FIG. 10 is a conceptual diagram for describing the watermark generation step 1403 shown in FIG. 4.

FIG. 11 illustrates an example of a target image in which geographic information is hidden according to a process of hiding geographic information in the image data shown in FIG. 3.

FIG. 12 is a diagram illustrating the target image illustrated in FIG. 11 divided into RGB (Red, Green, Blue) images.

FIG. 13A is a diagram illustrating a plurality of image blocks in which the target image illustrated in FIG. 11 is divided into 8 × 8 pixel units.

FIG. 13B is a diagram illustrating the image blocks shown in FIG. 13A divided into luminance (Y) components and color difference (Cb, Cr) components.

FIG. 14 is a diagram illustrating a plurality of image blocks in which an image of a color difference (Cb) component illustrated in FIG. 13B is divided into 8 × 8 pixel units.

FIG. 15 is a conceptual diagram illustrating a process of zigzag scanning after color difference (Cb) component values included in one of the plurality of image blocks shown in FIG. 14 are converted into color difference (Cb) frequency coefficient values by a discrete cosine transform. .

FIG. 16 is a diagram illustrating a plurality of image blocks in which an image of a color difference (Cr) component illustrated in FIG. 13B is divided into 8 × 8 pixel units.

FIG. 17 is a conceptual diagram illustrating a process of zigzag scanning after color difference (Cr) component values included in one of the plurality of image blocks shown in FIG. 16 are converted to color difference (Cr) frequency coefficient values by a discrete cosine transform. .

FIG. 18 illustrates an arrangement state after the color difference (Cb) frequency coefficient values shown in FIG. 15 are zigzag scanned.

FIG. 19 illustrates an arrangement state after the color difference (Cr) frequency coefficient values shown in FIG. 17 are zigzag scanned.

20 is a diagram illustrating an example of a quantization table used in the quantization step 1505 illustrated in FIG. 5.

FIG. 21 is a conceptual diagram illustrating a range of remaining values with respect to the adjusting step 1507 shown in FIG. 6.

22 is a diagram illustrating JPEG encoded images and images in which geographic information is concealed in a JPEG encoding process according to a geographic information concealment process according to the present invention.

FIG. 23 is a diagram illustrating images according to a result of executing various attack experiments on the image shown in FIG. 22B.

FIG. 24 is a table showing result values related to the attack experiment shown in FIG. 23.

FIG. 25 is a table showing values of execution results of various attack experiments on the image shown in FIG. 22D.

Claims (10)

Receiving GPS data; Extracting latitude and longitude values from the GPS data; Dividing a target image into which a watermark is to be inserted in units of 8x8 pixels, and setting a plurality of image blocks; Generating a watermark based on the number of the plurality of image blocks, the latitude value, and the longitude value; And In the process of encoding the target image data in JPEG (Joint Photographic Experts Group) encoding, by adjusting the size of the first or second color difference component of each of the plurality of image blocks according to the bit value of the watermark, geographic information is concealed Generating the generated image data, The bits of the watermark correspond to the plurality of image blocks, respectively. Generating the watermark, The latitude value and the longitude value of the latitude value and the longitude value are set based on Equation 1 and Equation 2 below, in order to set different degrees of importance for each digit of the latitude value and the longitude value. Calculating the number of repetitions for each digit; Generating a bit string by converting the latitude value and the longitude value of which the number of digits are increased by the number of repetitions into binary-coded decimal numbers (BCD); And Rearranging the bits of the bit string based on a predetermined key value defining a sequence of bits to generate a watermark. [Equation 1]

[Equation 2]

delete The method of claim 1, wherein the generating of the image data in which the geographic information is concealed comprises: Selecting one of the plurality of image blocks; Extracting luminance component values and first and second chrominance component values of the selected image block; Converting the luminance component values into luminance frequency coefficient values, the first color difference component values into first color difference frequency coefficient values, and the second color difference component values into second color difference frequency coefficient values, respectively, by a discrete cosine transform ; Zigzag scan and rearrange the luminance frequency coefficient values, the first color difference frequency coefficient values, and the second color difference frequency coefficient values, respectively; Quantizing the rearranged luminance frequency coefficient values and the first and second chrominance frequency coefficient values, respectively; Calculating a difference value between direct current (DC) component values of the quantized first and second chrominance frequency coefficient values; Quantized such that a difference value between direct current (DC) component values of the quantized first and second chrominance frequency coefficient values satisfies a condition set according to a corresponding bit value of the watermark corresponding to the selected image block Adjusting a DC component value of the first or second color difference frequency coefficient values; Repeating the selecting step and adjusting step until the DC component value of the quantized first or second chrominance frequency coefficient values of each of the entire image blocks is adjusted; And After the repetition step, entropy coding the quantized luminance frequency coefficient values and the first and second chrominance frequency coefficient values of which the DC component values are adjusted in all the image blocks. How to conceal. The method of claim 3, wherein adjusting the DC component values of the quantized first or second chrominance frequency coefficient values comprises: Divide the difference value between the DC component values of the quantized first and second chrominance frequency coefficient values by a set divisor value M (M = 4H, where H is an integer greater than 0) and divide the rest of the values. Obtaining; The first color difference frequency quantized such that when the residual value is greater than M / 2 and the corresponding bit value of the watermark corresponding to the selected image block is 1, the residual value is greater than 0 and less than M / 2. Decreasing the DC component value of the coefficient value or increasing the DC component value of the quantized second chrominance frequency coefficient value; And The quantized first chrominance frequency such that when the residual value is less than M / 2 and the corresponding bit value of the watermark corresponding to the selected image block is 0, the residual value is greater than M / 2 and less than M Increasing the DC component value of the coefficient values or decreasing the DC component value of the quantized second chrominance frequency coefficient values. The method of claim 4, wherein adjusting the DC component values of the quantized first or second chrominance frequency coefficient values comprises: When the remaining value is greater than 0 and less than M / 2, and the corresponding bit value of the watermark corresponding to the selected image block is 1, the DC component values of the quantized first or second chrominance frequency coefficient values are intact. Maintaining; And When the remaining value is larger than M / 2 and smaller than M and the corresponding bit value of the watermark corresponding to the selected image block is 0, the DC component value of the quantized first or second chrominance frequency coefficient values is determined. And concealing geographic information in the image data, the method further comprising maintaining the same. The method of claim 4, wherein adjusting the DC component values of the quantized first or second chrominance frequency coefficient values comprises: The remaining value is greater than 0 and less than M / 2 but not M / 4, and when the corresponding bit value of the watermark corresponding to the selected image block is 1, the remaining value is M / 4; Increasing or decreasing the DC component value of the first or second chrominance frequency coefficient values; Maintaining the DC component values of the quantized first or second chrominance frequency coefficient values when the remaining value is M / 4 and the corresponding bit value of the watermark corresponding to the selected image block is 1; The remaining value is greater than M / 2 and less than M but not 3M / 4, and when the corresponding bit value of the watermark corresponding to the selected image block is 0, the remaining value becomes 3M / 4, Increasing or decreasing the DC component value of the first or second chrominance frequency coefficient values; And Maintaining the DC component values of the quantized first or second chrominance frequency coefficient values when the remaining value is 3M / 4 and the corresponding bit value of the watermark corresponding to the selected image block is 0; A method of concealing geographic information in an image data further comprising. The method of claim 4, wherein adjusting the DC component values of the quantized first or second chrominance frequency coefficient values comprises: When the residual value is greater than -M / 2 and the corresponding bit value of the watermark corresponding to the selected image block is 1, the residual value is greater than -M and less than -M / 2; Decreasing the DC component value of one chrominance frequency coefficient values or increasing the DC component value of the quantized second chrominance frequency coefficient values; And The first quantized such that when the residual value is less than -M / 2 and the corresponding bit value of the watermark corresponding to the selected image block is zero, the residual value is greater than -M / 2 and less than zero. Increasing the DC component value of the chrominance frequency coefficient values or decreasing the DC component value of the quantized second chrominance frequency coefficient values. The method of claim 4, wherein adjusting the DC component values of the quantized first or second chrominance frequency coefficient values comprises: DC component values of the quantized first or second chrominance frequency coefficient values when the remaining value is greater than -M and less than -M / 2 and the corresponding bit value of the watermark corresponding to the selected image block is 1 Maintaining the same; When the remaining value is greater than -M / 2 and less than 0 and the corresponding bit value of the watermark corresponding to the selected image block is 0, the DC component value of the quantized first or second chrominance frequency coefficient values is determined. And concealing geographic information in the image data, the method further comprising maintaining the same. The method of claim 4, wherein adjusting the DC component values of the quantized first or second chrominance frequency coefficient values comprises: When the remaining value is greater than -M / 2 and less than 0 but not -M / 4 and the corresponding bit value of the watermark corresponding to the selected image block is 0, the remaining value becomes -M / 4, Increasing or decreasing the DC component value of the quantized first or second chrominance frequency coefficient values; Maintaining the DC component values of the quantized first or second chrominance frequency coefficient values when the remaining value is -M / 4 and the corresponding bit value of the watermark corresponding to the selected image block is 0 ; When the remaining value is greater than -M and less than -M / 2 but not -3M / 4, and the corresponding bit value of the watermark corresponding to the selected image block is 1, the remaining value becomes -3M / 4. Increasing or decreasing the DC component value of the quantized first or second chrominance frequency coefficient values; And Maintaining the DC component values of the quantized first or second chrominance frequency coefficient values when the remaining value is -3M / 4 and the corresponding bit value of the watermark corresponding to the selected image block is 1 How to conceal geographic information in the image data further comprising. Receiving GPS data; Extracting latitude and longitude values from the GPS data; Dividing the target image into which the watermark is to be inserted in units of 8x8 pixels, and setting a plurality of image blocks; Generating a watermark based on the number of the plurality of image blocks, the latitude value, and the longitude value; And In the process of JPEG encoding the target image data, by adjusting the size of the first or second color difference component of each of the plurality of image blocks, according to the bit value of the watermark, to generate the image data concealed geographic information A computer-readable recording medium storing a geographic information concealment program for executing a step, The bits of the watermark correspond to the plurality of image blocks, respectively. Generating the watermark, The latitude value and the longitude value of the latitude value and the longitude value are set based on Equation 1 and Equation 2 below, in order to set different degrees of importance for each digit of the latitude value and the longitude value. Calculating the number of repetitions for each digit; Generating a bit string by converting the latitude value and the longitude value of which the number of digits are increased by the number of repetitions into binary-coded decimal numbers (BCD); And And rearranging the bits of the bit string based on a predetermined key value defining a sequence of bits to generate a watermark. 17. A computer-readable recording medium having stored therein a geographic information hiding program. [Equation 1]

[Equation 2]

Images