JP2003153228A - Image encoding apparatus and image decoding apparatus - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an image encoding apparatus and an image decoding apparatus that can realize partially disclosed information a degree of scrambling of which can freely be controlled without increasing a data quantity after encryption. SOLUTION: The image encoding apparatus encrypts part of an encoded code stream in compliance with the JPEG-2000. The apparatus comprise, a means that uses a filter bank comprising a low pass filter and a high pass filter to apply wavelet transform to input image data a quantization means, a means for deploying a quantization coefficient into a bit plane from the MSB to the LSB, a means for applying arithmetic encoding to the bit plane by each unit of a prescribed entropy coding processing, a means for segmenting the generated arithmetic code to have a prescribed bit rate, a means for generating a header for the arithmetic code subjected to rate control, a means for encrypting the arithmetic code and a means for configuring a packet with the arithmetic code after encryption and the header.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image coding apparatus and an image decoding apparatus for partially encrypting a coded image bitstream according to JPEG-2000 (referred to as scrambling) and performing a cryptographic decoding process. The fields of application of the present invention are fields that require security, copyright protection, and privacy protection, and are extremely diverse. For example, in remote diagnosis of medical images, it is possible to scramble the images in advance in order to protect the privacy of the patient. Also, if an image having a copyright is scrambled, the original image cannot be reproduced unless the receiving side has an encryption key, which is an effective means for copy protection.

[0002]

2. Description of the Related Art There are three means for copyright protection: an information semi-disclosure method in which data is scrambled mainly by digital watermarking, complete image encryption processing, or partial image encryption processing. Among them, digital watermarking is a technique for suppressing unauthorized use by embedding copyright information in an image. Since the original image quality is lost due to the embedding operation, it is not suitable for cases where high quality is required or applications where lossless encoding is premised. On the other hand, the complete image encryption method is considered to be the most robust method,
The complete encryption process has a drawback that the ease of handling as a medium is lost.

Therefore, as a new way of thinking about copyright protection, JP, which is an international standard for still images, which is now widely used.
NTT is an information semi-disclosure method that scrambles the encoded code stream of an EG image using partial encryption to deteriorate the quality to the extent that the outline of the image can be recognized.
Proposed by the research institute (Fujii, Abe, Nishihara, Kushima, "Information Semi-Disclosure Method", NTT R & D, Vol.47, No.6, pp.7)
05-710, Jun. 1998). In this method, the image provider
The scrambled image is distributed through the network or CD-ROM, and the user obtains the encryption key from the provider and uses it to descramble the image.

[0004]

The scrambling method for the JPEG coded code stream has a problem that the amount of data after scrambling increases, and there is a drawback that the compression rate becomes poor.

The present invention has been made in view of such circumstances, and an object thereof is to realize an information semi-disclosure means capable of freely controlling the degree of scrambling without increasing the amount of data after scrambling. The present invention provides an image encoding device and an image decoding device that can achieve such an object.

[0006]

In order to achieve the above object, an image coding apparatus according to the present invention comprises a coding means for coding an input image signal in accordance with a predetermined image processing standard, and the coded coding. A means for separating / extracting an encoded code stream to be encrypted from the code stream, an encryption means for encrypting the extracted encoded code stream, and an encoding based on the encrypted data. Means for writing back to the same part of the codestream.

Further, in the present invention, it is preferable that the coded code stream to be encrypted is an arithmetic code which is EBCOT output data defined by a predetermined image processing standard.

Further, in the present invention, it is preferable that the coded code stream to be encrypted is coded data of a packet body defined by a predetermined image processing standard, or, of the packet bodies. , Encoded data of a predetermined code block.

Further, in the present invention, it is preferable that the encryption means is a block encryption means using common key encryption.

Further, in the present invention, it is preferable that the coded code stream to be encrypted is coded data of a code block corresponding to a predetermined image area.

Further, in the present invention, it is preferable that the coded code stream to be encrypted is coded data of a predetermined coding pass in a code block corresponding to a predetermined image area.

Further, in the present invention, it is preferable that the coded code stream to be encrypted is coded data of a packet body existing in a predetermined layer.

Further, in the present invention, it is preferable that the coded code stream to be encrypted is coded data of a code block of a packet body existing in a predetermined layer.

Further, in the present invention, it is preferable that the coded code stream to be encrypted is coded data of a predetermined coding path among code blocks in a packet body existing in a predetermined layer.

Also, the image coding apparatus of the present invention comprises a transforming means for wavelet transforming the input image signal, a quantizing means for quantizing the transform result of the wavelet transforming means, and a quantum obtained by the quantizing means. Means for expanding the coding coefficient into a bit plane from the most significant bit to the least significant bit; means for performing an arithmetic coding process on the bit plane for each predetermined entropy coding processing unit; and the arithmetic coding process Means for clipping the arithmetic code generated by the above so as to have a predetermined bit rate,
It has means for creating a header in the arithmetic code whose bit rate is controlled, means for encrypting the arithmetic code, and packet generation means for generating a packet by the encrypted arithmetic code and the header.

Also, the image decoding apparatus of the present invention receives the partially encrypted coded code stream, separates and extracts the coded code stream to be partially encrypted, and the extracted part. Decryption means for decrypting encrypted encoded data using a common key, unencrypted encoded data in the encoded code stream, and encoded decrypted by the decryption means Means for arithmetically decoding the data and.

Further, the image decoding apparatus of the present invention comprises a packet separating means for separating a packet header and a packet body from an input partially encrypted code stream, and a code block or a coding path in the packet body. Means for truncating a part of the coded data of, a means for rewriting the code block and coding path information of the coded code stream after the truncation processing in a packet header, and a new packet from the rewritten packet header and packet body And generating means for generating.

[0018]

1 is a block diagram showing the overall configuration of an image coding apparatus and an image decoding apparatus according to the present invention. FIG. 1A shows the configuration of the image encoding device, and FIG. 1B shows the configuration of the image decoding device. As shown in the figure, the image encoding device is composed of an image encoding unit 1, an encryption target selecting / extracting unit 2 and a partial encryption unit 3, and the image decoding device is a partial decryption unit 4 and an image. It is composed of a decoding unit 5.

The present invention realizes an information semi-disclosure means specialized in the JPEG-2000 system, which has been standardized in January 2001 as the next-generation international standard for still images. Also in this case, of the above two conditions, both the condition 1 can control the degree of scrambling, and the condition 2 can satisfy both of the fact that the amount of data after scrambling does not increase.

Next, the operation of the image coding apparatus and the image decoding apparatus according to the present invention will be described with reference to FIG.
First, in the image encoding device, the image encoding unit 1 uses the JP
The input image 100 is encoded according to an EG-2000 compliant algorithm to generate an encoded code stream 101. Generated JPEG-2000 encoded codestream 10
In the encryption target selection / extraction unit 2, 1 is separated into an encryption target part and other parts. The partial encryption unit 3 inputs the separated encoded codestream 102 and encrypts only the encoded codestream to be encrypted. The same unit 3 also performs a process of writing back the encrypted code stream that has been encrypted to the part before encryption. As a result, the partial encryption process can be realized without increasing or decreasing the data amount of the entire encoded code stream. The encoded codestream 103 partially encrypted by the partial encryption unit 3 is transmitted.

Next, in the image decoding apparatus, the partial descrambling unit 4 decodes the codestream of the encrypted image portion with respect to the encoded codestream 103 of the partial encryption. Then, the decrypted image portion is written back to the original portion, and the deciphered encoded code stream 104 is generated. The image decoding unit 5 performs image decoding processing on the coded code stream 104 that has been decrypted in accordance with a JPEG-2000 compliant algorithm, and reproduces the image data 105. Next, the image coding apparatus of the present invention will be described in more detail for each embodiment.

First Embodiment FIG. 2 is a circuit diagram showing a first embodiment of an image coding apparatus according to the present invention. As shown in the figure, the image coding apparatus according to the present embodiment includes a wavelet transform unit 6, a quantization unit 7,
Code blocking unit 8, coefficient bit modeling unit 9, arithmetic coding unit 10, rate control unit 12, partial encryption target selection / extraction unit 13, encryption unit 14, and packet generation unit 15.
It is composed by. In this, the code blocking unit 8, the coefficient bit modeling unit 9, and the arithmetic coding unit 1
0 constitutes an EBCOT encoding unit 11.

The configuration and operation of each part of the image coding apparatus of this embodiment will be described below. The wavelet transform is usually realized by a filter bank composed of a low pass filter and a high pass filter. Therefore, since the digital filter usually has an impulse response (filter coefficient) of a plurality of tap lengths, it is necessary to buffer an input image that can be filtered in advance.
Note that FIG. 2 does not show an input buffer that buffers the input image data.

The wavelet transform unit 6 to which the minimum image data 100 necessary for filtering has been input as described above performs filtering for wavelet transform, and the wavelet transform coefficient 106 is generated here. The wavelet transform coefficient 106 is subsequently quantized by the quantizer 7, and the quantized coefficient 107 is output. As the quantizing means, the normal wavelet transform coefficient 106
Scalar quantization, which divides by the quantization step size, is common. Incidentally, the wavelet transform usually takes a means for repeatedly transforming the low frequency components as shown in FIG. 3, because most of the image energy is concentrated in the low frequency components.

In the case of FIG. 3, the number of wavelet transform levels is 2, and as a result, a total of 7 subbands are generated. Incidentally, in the ordinary wavelet transform, a means for recursively filtering only low frequency components is general.

Next, the code blocking unit 8 divides the quantized coefficients 107 in the sub-band divided by the wavelet transform into code block units which are JPEG-2000 coding units. FIG. 4 illustrates this code blocking, and it can be seen that a code block of 64 × 64 size is generated for each subband shown in FIG. In the JPEG-2000 standard, the code block size is represented by a power of 2 both horizontally and vertically, and 32x32 or 64x64 is normally used.

The quantized coefficient 108 for each code block is modeled in bit plane units in the coefficient bit modeling section 9 and in the coefficient bit modeling section 9 defined by JPEG-2000. Here, the concept of the bit plane will be described with reference to FIG. The left diagram of FIG. 5 assumes a quantized coefficient consisting of a total of 16 coefficients, four vertically and four horizontally. Of these 16 coefficients, the one with the largest absolute value is 13, which is 1101 when expressed in binary. Therefore, the bit plane of the absolute value on the right side of the figure is composed of four planes. It is clear that the elements of each bit plane take a number of 0 or 1. On the other hand, the sign is -6, which is the only negative number, and the others are 0 or positive numbers. Therefore, the code plane is as shown on the right in the figure.

EBCOT (Embedd) defined by the JPEG-2000 standard
Entropy coding process called ed Coding with Optimized Truncation (Reference: ISI / IEC FDIS 15444-
1, JPEG-2000 Part-1 FDIS, 18 August 2000) is realized by the coefficient bit modeling unit 9 and an arithmetic coding unit 10 described later. The processing unit of EBCOT is the above code block.

The code block is independently coded for each bit plane in the MSB to LSB direction. The quantized coefficient value is represented by an n-bit signed binary number, and bit 0
To bit (n-2) represent each bit from LSB to MSB. The remaining 1 bit is a code. The coding of the code block is performed by the following three types of coding paths in order from the MSB side bit plane. (1) Significance Propagation Pass (2) Magnitude Refinement Pass (3) Cleanup Pass

The order in which the above three coding passes are used is shown in FIG. First, the bit plane (n
-2), that is, the MSB is encoded by the Cleanup Pass. Then, toward the LSB side, the coding of each bit plane is performed in three coding passes.
pagation Pass, Magnitude Refinement Pass, Cleanu
It is performed using the order of p Pass. However, in reality MSB
In the header, the bit plane from which the 1 is first output is written in the header, and the bit plane in which all data is 0 is not encoded at first.

The coding is repeated using three types of coding passes in this order, and the coding is cut off up to an arbitrary coding pass of an arbitrary bit plane, thereby taking a trade-off between the code amount and the image quality (that is, Rate control).

Next, coefficient scanning (scanning) will be described with reference to FIG. The code block is divided into stripes for each coefficient of height 4. The width of stripe is equal to the width of the code block. The scan order is the order in which all the coefficients in a code block are traced, from the upper stripe to the lower stripe in the code block, and from the left column to the right column in the stripe. The order is from top to bottom in the sequence. All coefficients of the code block in each coding pass are processed in this scan order.

Next, the above three coding passes will be described. (1) Significance Propagation Pass Significance Propagation Pass that encodes a bit plane
In tion Pass, at least one coefficient near 8 is signi
Arithmetically encode the bit-plane values of non-significant coefficients such as ficant. When the value of the coded bit plane is 1, whether the code is + or − is continuously subjected to arithmetic coding.

Here, the term “significance” will be described. Significance is the state that the encoder has for each coefficient, and the initial value of significance is non-significant.
Is 0, which is the signific when 1 is encoded with the coefficient
It changes to 1, which represents ant, and always remains 1. Therefore, the significance can be said to be a flag indicating whether or not the significant digit information has already been encoded.

(2) Magnitude Refinement Pass Magnitude Refinement Pas for encoding the bit plane
s is the Significance Pr that encodes the bitplane
The value of the bit plane of the significant coefficient which is not encoded by opagation Pass is arithmetically encoded.

(3) Cleanup Pass In the Cleanup Pass for encoding the bit plane, the Significance Propagation Pass for encoding the bit plane is used.
The value of the bit plane of the non-significant coefficient that is not coded by is arithmetically coded. When the value of the encoded bit plane is 1, whether the sign is + or − is continuously arithmetically encoded.

In the arithmetic coding with the above three coding passes, ZC (Zero Coding) and R are used depending on the case.
LC (Run-Length Coding), SC (Sign Coding), MR
(Magnitude Refinement) is used properly. here,
JPEG-2000 uses an arithmetic code called MQ coding. MQ encoding is based on JBIG2 (reference: ISO / IEC FDIS
14492, “Lossy / Lossless Coding of Bi-level Image
s ", March 2000), which is a learning type binary arithmetic code. In JPEG-2000, there are a total of 19 types of contexts in all coding passes.

The configuration and function of the coefficient bit modeling unit 9 and the arithmetic coding unit 10 provided in conformity with the JPEG-2000 standard in the image coding apparatus shown in FIG. 2 have been described above. Next, the latter half of the image coding apparatus shown in FIG. 2 will be described.

First, the rate control unit 12 controls the code amount so as to approach the target bit rate or compression ratio while counting the code amount of the arithmetic code 110 output from the arithmetic coding unit 10. Specifically, it can be realized by truncating part or all of the coding path for each code block. Up to here JPEG-200
0 It is an encoding process based on the standard.

The coded code stream 111 for each code block after the rate control is input to the partial encryption target selection / extraction unit 13. The partial encryption target selecting / extracting unit 13 receives the information 116 indicating which part of the JPEG-2000 encoded code stream is to be encrypted, and the encoded code stream 112 to be encrypted and the other encodings. The code stream 115 is divided.

Encoded code stream 11 to be encrypted
2 is encrypted by a means using a common key by an encryption means described later, and the encrypted code stream 11 is encrypted.
4 is output from the encryption unit 14. Finally, the packet generator 15 packetizes the encrypted coded code stream 114 and the unprocessed coded code stream 115, respectively, and outputs the packetized coded code stream 117. What is this packet?
It consists of a packet header and packet body specified by JPEG-2000.

In the packet header, the number of coding passes in the code block, the data length of the code stream for each coding pass, etc. are described according to the JPEG-2000 standard.

Next, the encryption described in the present invention will be described. Cryptographic technology is a reversible data conversion technology, which is the source of communication messages (called plaintext).
Is a technology that converts data into completely different data (called ciphertext) from the previous one according to a certain rule to make it useless to a third party who does not know the conversion rule. The process of converting plaintext into ciphertext is called encryption, and conversely the process of converting ciphertext into plaintext is called decryption. The reversible data conversion technique requires at least a one-to-one correspondence between plaintext and ciphertext. As a possible variation for making a one-to-one correspondence like this, when the plaintext is N-bit data, 2 ^N ! There can be street patterns.

It can be said that the cryptographic technique prepares a table of all possible rearrangement patterns according to the size of the plaintext in this way, selects one pattern in the table and outputs the corresponding ciphertext. At that time, instead of continuing to use the same pattern, a parameter for dynamically deciding which pattern to use as needed is used. This parameter is called "key". Even if the same algorithm is used, the ciphertext for the same plaintext can be made different by changing the key.

However, it is not realistic to prepare a conversion table that exhausts all variations for plain texts of a certain size or more. Therefore, the actual encryption method focuses on how to efficiently realize a subset of all conversion patterns, and tries to realize conversion that is hard to guess without using such a large amount of data. is there. The actual encryption methods are roughly classified into common key encryption and public key encryption.

The common key cryptography is a system in which the same key is used for encryption and decryption, and the security of data cannot be maintained unless this key is kept secret. In principle, as a magic number (key) for two-way transformation, which is a one-to-one reversible transformation that maps the existence space of plaintext blocks (possible values of bit strings) and the existence space of ciphertext blocks. The same value is used. Usually, the decryption process is performed in the reverse order of the encryption process. There are two types of ciphers, block ciphers and stream ciphers, depending on how to operate plaintext data. In the present invention, a block encryption means, which is commonly used and can be encrypted in block units, is used. The reason is that first, the arithmetic code is a unit of a fixed block length, and therefore the compatibility is good, and that the amount of data does not change before and after encryption.

In the block cipher, a plaintext is divided into blocks each having a fixed block length, an encryption process is performed by using a common key, and a decryption is also performed by using the common key in the same block unit. If the plaintext is less than the block length, padding (replenishment processing) is required. The main common block encryption method is the patent-free Blowfish method. This is a 64-bit block cipher system developed by Bruce Schneier and released as free software. The key length is variable and can be 32 bits to 448 bits.

If this block encryption is used, the data string of the arithmetic code, which is a set of 0s and 1s, may be divided into fixed lengths, and those fixed length blocks may be encrypted. It has the advantage that the quantity does not increase. This satisfies the object of the present invention that the amount of data does not increase. The EBCOT used in JPEG-2000 outputs data in 1-byte units (has byte alignment), so it has the feature of being very compatible with this block encryption.

Second Embodiment This embodiment is an embodiment relating to the partial encryption target described in the above-mentioned first embodiment of the present invention. The image coding apparatus according to this embodiment has substantially the same configuration as that shown in FIG.

In the above-described first embodiment of the present invention, JP
The EG-2000 encoded codestream is composed of units called packets, and each packet consists of a packet header and a packet body. In this packet header,
The JPEG-2000 standard describes the number of coding passes in a code block and the data length of the code stream for each coding pass. Therefore, the image cannot be restored unless the packet header is decoded correctly.

Therefore, in this embodiment, only the packet body is to be encrypted. As a result, even the receiving side without the encryption / decryption function can decrypt the JPEG-2000 encoded codestream. However, the decoded image deteriorates.

In addition to the present embodiment in which the packet body is the object of encryption, the code block that is a constituent element of the packet body and the coding path that is a constituent element of the code block can also be the object of encryption. Is.

As described above, the coded code stream of the JPEG-2000 standard is composed of a unit called a packet, and the actual coded code stream is recorded in the packet body of the packet. Therefore, it is essential to correctly detect and read this packet body.

In the JPEG-2000 standard, SOP (Start of Packet), so that decoding can be performed to some extent even if the coded code stream is broken for some reason,
Two markers of EPH (End of Packet Header) are prepared. The JPEG-2000 standard has a mechanism that can detect as many packets as possible by detecting these two markers in the encoded code stream. However, in the JPEG-2000 standard, this SOP and EPH are optional, and the encoder side selects whether or not to include the SOP and EPH in the encoded code stream.

FIG. 8 shows a packet header and a packet body in the case where the SOP and EPH are included in the encoded code stream and in the case where the SOP and EPH are not included in the encoded code stream, and a positional relationship between them when the SOP and the EPH are included.

FIG. 8A shows the case where the SOP and the EPH are included. The first packet is composed of the SOP, the packet header EPH, and the packet body in this order. On the other hand, FIG. 8B shows a case where SOP and EPH are not included, and the data is arranged in the order of the packet header and the packet body.

JPEG-2000 shows whether the SOP and EPH are included in the coded code stream.
This is the Scod parameter in the COD marker defined in the standard. It is defined in the JPEG-2000 standard.

FIG. 9 is a flowchart showing a series of processes. The encryption process will be sequentially described below with reference to FIG. First, the encryption target is determined in step S10. For example, it is determined whether an entire packet body, a code block, or a coding pass (step S00). Next, the Scod parameter in the above-mentioned COD marker is read, and it is judged from this that whether SOP and EPH are included (step S20).

As a result of the determination, the process is branched depending on whether the SOP and EPH are included or not (step S30). If the SOP and EPH do not exist, the process proceeds to step 1 (step S40), and if they do exist, the process proceeds to step 2 (step S50). Hereinafter, each of these two processes will be described.

First, the processing 1 when it does not exist will be described with reference to the flowchart shown in FIG. As already described in FIG. 8B, the packet is composed of the packet header and the packet body. Therefore, in process 1, the packet header is first detected (step S60). As mentioned above, the packet header contains JPEG-2
The 000 standard describes the number of coding passes in a code block and the data length of the code stream for each coding pass. Therefore, in order to detect the packet body located after the packet header, it is necessary to correctly decode the packet header (step S70).

Next, it is determined whether the packet body is an encryption target (step S80). If the packet body is an encryption target, the process proceeds to step S110. On the other hand, if the packet is not the encryption target, the process is repeated until the target packet is found. At that time, since the data length of the subsequent packet body can be known by decoding the packet header, the end of the packet body can be immediately detected (step S90). Therefore, the next packet header may be similarly detected and the decoding process may be performed.

Next, the operation when the packet body is an encryption target will be described. First, the packet body is detected (step S110). Then, the detected packet body is encrypted (step S120). Furthermore, the encrypted packet body is written back to the original encoded code stream (step S130), and the process ends.

Next, processing 2 shown in FIG. 9 (step S5
0) will be described. The process 2 is, that is, SOP.
And the EPH marker is included in the encoded code stream. Hereinafter, a specific operation will be described with reference to FIG. First, from the encoded code stream, the number of layers (Layer, the number of layers when packetizing), the number of components (the number of Components, RGB
If so, three components), the number of decompositions (the number of Decomposition, that is, the number of wavelet divisions),
Each value of Precinct-size (size of precinct of JPEG-2000 standard) and Progression-Order (indicating any of five progressions defined by JPEG-2000 standard) is detected (step S150).

Specifically, Progression-Order and Layer
The number is specified by a parameter called SGcod in the COD marker. The number of Decomposition and Precinct-size are specified by the parameter SPcod. Furthermore, the number of Components is specified by the parameter Csiz of SIZ marker. Each of these parameters is detected (step S160). Then, by reading the detected parameters (step S160), the numerical values of the respective parameters are recognized.

Next, by reading the parameter Nsop defined in the SOP (step S170),
The sequence number of the packet is detected. The sequence number is a numerical value of a serial number indicating the number of packets counted from the beginning of the encoded code stream,
By detecting this sequence number, it can be determined whether or not the current packet is an encryption target. Therefore, if it is not the packet to be encrypted, the encoded code stream is read until the next SOP marker is found (step S190).

On the other hand, the EPH marker is detected (step S180). If the packet is an encryption target (step S200), subsequent processing (step S21)
Proceed to 0). As described above, in this processing, it is not necessary to decode the packet header, and the load on the decoder is reduced.

Next, the processing of the packet to be encrypted will be described. First, the packet body following the EPH marker is detected (step S210). Then, the detected packet body is encrypted (step S22).
0). The encrypted packet body is written back to the encoded code stream again (step S130), and the process ends.

Encryption Processing for Code Stream (1) Next, the extraction processing of the encoded code stream to be encrypted will be described with reference to FIG. FIG. 12 (A)
When the encryption processing is performed on the image corresponding to the shaded area (lower right part) of the sub-band, it exists in each sub-band area formed by wavelet division as shown in FIG. The coded code stream of the code block may be encrypted. Note that, as an example, FIG. 13 shows an example of an encrypted image when the central area of the screen is encrypted. When decryption is performed without a key, it can be seen that the image of the encrypted portion is deteriorated, as shown in FIG.

It is also possible to limit the encryption target to a specific coding pass in the code block. In that case, there is an effect that the degree of freedom of the encryption target is increased and the image quality of the target area image can be variously changed.

Encryption Process for Code Stream (Part 2) Next, another example of the extraction process of the encoded code stream to be encrypted will be described. FIG. 14 shows an example of this processing. In this processing example, an encoded code stream for a subband of a certain resolution level formed by wavelet division is encrypted. Figure 1
In the left figure of 4, the encryption is performed on the coded code stream of the packet body existing in the resolution level of level 2 (HL2, LH2, HH2). Similarly, in the right diagram of FIG. 14, the encrypted code stream of the packet body existing in the resolution level of level 1 (HL1, LH1, HH1) is encrypted.

Here, the relationship between the resolution level and the packet will be briefly described. The JPEG-2000 standard uses the aforementioned method called Progression-Order to determine the order in which the coded code streams are arranged from the beginning. For example, from the image with the smaller resolution to the image with the gradually larger resolution, from the image with the lower image quality to the image with the higher image quality,
Or, in an RGB image, there is the order of R->G-> B components.

FIG. 15 is a conceptual diagram in which the coded code streams are arranged in order from the smallest resolution to the largest resolution in the above Progression-Order. This figure is a diagram when layering described later is performed, and in the case of one layer, only Layer-0 needs to be viewed. Layer-0 is the lowest Pa
cket-1 (resolution level 0, the corresponding subband is LL-0
1) to Packet-1 (at resolution level 1, corresponding subbands are HL-1, LH-1, and HH-1), Packet-2
(At resolution level 2, corresponding sub-bands are HL-2, LH-
2, HH-2) and Packet-3 (at resolution level 3, corresponding subbands are HL-3, LH-3, and HH-3).

The two figures shown in FIG. 16 are images when all the packet bodies of level 1 only and level 3 only are encrypted and decrypted by a decryptor having no common key. Similarly, FIG. 17 is an image when the packet bodies of level 4 and level 5 are encrypted. From these, it can be seen that the purpose of so-called "information semi-disclosure" is achieved in which the overall image quality is deteriorated while maintaining the general characteristics of the image.

[0074] It is also possible to limit the subject of encryption encryption process for the code block to a particular code block in the packet body. In that case, there is an effect that the degree of freedom of the encryption target is increased and only the specific area of the target resolution level can be encrypted. Figure 1
8 is a flowchart showing the specific processing procedure. Hereinafter, with reference to FIG. 18, description will be given of a process when the target of the encryption process is limited to a specific code block in the packet body.

First, it is determined whether the entire packet body is to be encrypted (step S310). When the entire packet body is the encryption target, as described above, the processes of steps S110 to S130 in the flowchart shown in FIG. 10 are performed. On the other hand, if not, the next code block in the packet body is detected (step S330), and it is determined whether the detected code block is an encryption target (step S340). This process is repeated to find the code block to be encrypted. When the target code block is detected, the process proceeds to step S350.

Subsequently, it is determined whether the entire code block is the encryption target (step S350). If the entire code block is the encryption target, the process proceeds to step S360 shown in FIG. On the other hand, if not, the next coding path in the code block is detected (step S370), and it is determined whether or not the detected coding path is an encryption target (step S370).
380). As soon as the coding path to be encrypted is detected, the process proceeds to step S390 shown in FIG.

FIG. 19 illustrates the process of step S360, which is a process of encrypting a target code block. First, the code block to be encrypted is encrypted (step S410), the encrypted code block is written back to the original encoded code stream (step S420), and the process ends.

FIG. 20 illustrates the process of step S390, which is a process for encrypting a target coding path. First, the coding path to be encrypted is encrypted (step S440), the encrypted coding path is written back to the original coded code stream (step S450), and the process ends.

Third Embodiment FIG. 21 is a circuit diagram showing a second embodiment of the image coding apparatus according to the present invention. In comparison with the first embodiment of the present invention shown in FIG. 2, in the image encoding device of the present embodiment, the partial encryption target selecting / extracting unit 13 of the first embodiment is replaced with the layering unit 16. It has been changed.

First, the packet layering process will be described with reference to FIG. When having multiple layers, Progression-Or in which the coded code streams are arranged in order from the smallest resolution to the largest resolution.
With der, layers from Layer-0 in the MSB direction to Layer-2 in the LSB can be created. Further, a packet is generated for each layer by the same means.

FIG. 22 illustrates the relationship between actual code blocks, coding paths, and layered packets. FIG. 15 and FIG. 22 correspond to each other. 22C is Cleanup-Pass, S is Signif
icance Propagation Pass, M is Magnitude Refinement
Shows each Pass. These coding paths have already been described.

FIG. 22 shows that there are a total of four code blocks from code block 0 to code block 3, and the coding paths forming each code block are layered. In Packet-0 of Layer-0, there are four coding paths C, S, M, and C for code block 0, and code block 1 of the same packet.
For, there is only C. There is no coding path in code block 1.

When the coded code stream of the packet body existing in a specific layer is encrypted,
As shown in FIG. 23, the entire coding path in the area surrounded by halftone lines is encrypted.

On the other hand, when the coded code stream of the code block in a certain packet is encrypted, for example, as shown in FIG. 24, the entire coding path in the area surrounded by the right half-tone dot mesh is encrypted. (In this figure, all coding passes of code block 3 are encrypted).

Further, when the coded code stream of a certain coding path within a certain code block is encrypted, for example, the coding path of the area surrounded by the mesh line at the right end of FIG. 25 is encrypted (this In the figure, only the coding pass C of the code block 3 is encrypted).

Fourth Embodiment FIG. 26 is a circuit diagram showing a third embodiment of the present invention, which is a circuit diagram showing one structural example of the image decoding apparatus of the present invention. As shown in the figure, the image decoding device of the present embodiment is
Packet decryption unit 21, partial encryption target selection / extraction unit 1
3, an encryption / decryption unit 22, a calculation decryption unit 23, a bit modeling decryption unit 24, a code block restoration unit 25, an inverse quantization unit 26, and a wavelet inverse transformation unit 27.

The image decoding apparatus according to this embodiment is shown in FIG.
It is a specific example of the image decoding apparatus of the present invention shown in (B). As shown in FIG. 1 (B), a partial descrambling unit 4 is provided in front of the image decoding unit 5 of the normal JPEG-2000 standard. The encrypted part is separated from the encrypted part and the other part is passed to the JPEG-2000 decoding section 5. However, in FIG. 1B, the partial decryption unit 4 is an example independent from the image decryption unit 5 of JPEG-2000, but in the image decryption apparatus of the present embodiment, the JPEG-2000 decryption unit is used. The function is included in. The image decoding apparatus according to the present embodiment has a configuration that forms a pair with the image encoding apparatus shown in FIG.

The operation of the image decoding apparatus of this embodiment will be described below with reference to FIG. First, the packetized encoded code stream 117 is transmitted to the packet header 21 and the packet body 128 in the packet decoding unit 21.
And separated. Next, based on the information 116 to be partially encrypted, the encoded code stream 114 that has been encrypted and the other encoded code streams 115 are separated.

Encrypted coded code stream 11
4 is decrypted by the encryption / decryption unit 22 using the common key 113. Both the coded code stream 112 after decoding and the non-encrypted normal coded code stream 115 are decoded by the arithmetic decoding unit 23 by the arithmetic decoding means defined in the JPEG-2000 standard. To be done. All subsequent processing may use the same means as defined in the JPEG-2000 standard.

Fifth Embodiment FIG. 27 is a circuit diagram showing a fourth embodiment of the present invention, which relates to a bit rate conversion circuit for bit rate converting an encoded code stream which has been encrypted. As shown in the figure, the bit rate conversion circuit according to the present embodiment includes a packet decoding unit 17, a lower layer truncation unit 18,
Packet header rewriting unit 19 and packet generation unit 20
It is composed by.

The operation of the bit rate conversion circuit of this embodiment will be described below with reference to FIG. The packetized encrypted code stream 117 and the encryption target information 113 are input, and the packet decryption unit
The encrypted packet and the other packet 122 are separated. Next, the lower layer truncation unit 18 determines which layer should be truncated based on the information of the designated bit rate 123. As a result, the coded code stream 126 of the packet body of the layer other than the truncated layer
Then, the original code block and coding path information 124 is output from the same section 18.

The original code block and coding path information 124 is used by the packet header rewriting unit 19 as a new code block and coding path information 125.
Can be rewritten as As a result, in the packet generator 20, a new packet header 125 and packet body 1
26 and a new packet 127 is output.

FIG. 28 shows an example in which this embodiment is applied to a layered coded code stream. In this application example, the bit rate including the coding paths in all the code blocks of the three layers is 1.0 bpp, and the bit rate including the coding paths in all the code blocks of the upper two layers is
It shows that the bit rate is 0.75bpp, and the bit rate including the coding path in the code block of only the highest layer is 0.50bpp.

Therefore, when the coding path in the code block of the uppermost layer which is shaded from this is encrypted, if the coding path in the code block of the lower layer is truncated, the target bit rate of An encoded code stream that has been encrypted can be created.

[0095]

As described above, according to the image coding apparatus and the image decoding apparatus of the present invention, the coded code stream compliant with the JPEG-2000 standard can be encrypted and the information semi-disclosure method can be adopted. Therefore, there is an effect that the contents of the image can be transmitted to the recipient who does not have the key.
Furthermore, the image quality can be controlled according to the application and the transmission partner. Of course, the receiver who has the key also has the effect of being able to restore the exact same image as the sender. Further, according to the present invention, since the strength of encryption can be realized by any means of resolution level of wavelet transform, layer, and depth of bit plane, there is also an effect that the encryption can be performed with extremely high flexibility. Furthermore, by applying the block encryption means to the data string of the arithmetic code divided by the fixed block length, there is an effect that the data amount does not increase even after scrambling. This is extremely important for the JPEG-2000 method, which has the advantage of high compression ratio.

[Brief description of drawings]

FIG. 1 is a block diagram showing an image encoding device and an image decoding device according to the present invention.

FIG. 2 is a circuit diagram showing a first embodiment of an image encoding device according to the present invention.

FIG. 3 is a diagram showing wavelet division with a resolution level of 2.

FIG. 4 is a diagram showing an example of a 64x64 code block in a wavelet division having a resolution level of 2.

FIG. 5 is a diagram showing a concept of a bit plane.

FIG. 6 is a diagram showing an example of three coding paths.

FIG. 7 is a diagram showing a scanning method in a code block.

FIG. 8 is a diagram showing a packet configuration when SOP and EPH markers are included and when they are not included.

FIG. 9 is a flowchart showing a branch of encryption processing depending on the presence / absence of SOP and EPH markers.

FIG. 10 is a flowchart showing an encryption process when there is no SOP and EPH marker.

FIG. 11 is a flowchart showing an encryption process when SOP and EPH markers are present.

FIG. 12 is a diagram showing an encryption process of a predetermined image area.

FIG. 13 is a diagram showing a decrypted image obtained by a normal decryption process after encryption of a predetermined image area.

FIG. 14 is a diagram showing a case where resolution levels are 1 and 2.

FIG. 15 is a diagram showing a structure of a layered packet.

FIG. 16 is a diagram showing a decrypted image obtained by a normal decryption process after an encryption process of a predetermined level.

FIG. 17 is a diagram showing a decrypted image obtained by a normal decryption process after the level 4 and level 5 encryption processes.

FIG. 18 is a flowchart showing a process in which an encryption target is a packet body or lower.

FIG. 19 is a flowchart showing a code block encryption process.

FIG. 20 is a flowchart showing a coding path encryption process.

FIG. 21 is a circuit diagram showing another embodiment of an image encoding device that performs partial encryption processing.

FIG. 22 is a diagram showing a relationship among code blocks, layers, packets, and coding paths.

[Fig. 23] Fig. 23 is a diagram illustrating a case where all coding paths in a packet body are encryption targets.

[Fig. 24] Fig. 24 is a diagram illustrating a case where all coding paths in a code block in a packet body are encryption targets.

[Fig. 25] Fig. 25 is a diagram illustrating a case where a certain coding path in a certain code block in a packet body is an encryption target.

[Fig. 26] Fig. 26 is a diagram illustrating the configuration of an image decoding device that performs partial encryption processing.

[Fig. 27] Fig. 27 is a diagram illustrating a bit rate conversion circuit of an encrypted encoded code stream.

FIG. 28 is a diagram of bit rate conversion by truncation of layered packets.

[Explanation of symbols]

1 ... Image coding unit, 2 ... Encryption target selection / extraction unit, 3 ...
Partial encryption unit, 4 ... Partial deciphering unit, ... Image decoding unit,
6 ... Wavelet transform unit, 7 ... Quantization unit, 8 ... Code blocking unit, 9 ... Coefficient bit modeling unit, 10 ... Arithmetic coding unit, 11 ... EBCOT coding unit, 12 ... Rate control unit, 13 ... Partial encryption Encryption target selection / extraction unit, 14 ... encryption unit, 15 ... packet generation unit, 16 ... layering unit, 17 ...
Packet decoding unit, 18 ... Lower layer truncation unit, 19 ...
Packet header rewriting unit, 20 ... Packet generating unit, 2
DESCRIPTION OF SYMBOLS 1 ... Packet decoding part, 22 ... Encryption decoding part, 23 ... Arithmetic decoding part, 24 ... Bit plane decoding part, 25 ... EBCOT
Decoding unit, 26 ... Dequantization unit, 27 ... Wavelet inverse transformation unit, 100 ... Input image, 101 ... Encoding code stream of JPEG-2000 standard, 102 ... Encoding object and encoding code separated into other parts Stream, 103 ... Partially encrypted coded code stream, 104 ... Encrypted / decoded coded code stream, 105 ... Decoded image, 106 ... Wavelet transform coefficient, 107 ... Quantized coefficient, 108 ... Code block Quantization coefficient, 1
09 ... Bit modeled quantized coefficient, 110 ... Arithmetic codeword, 111 ... Rate-controlled arithmetic codeword, 112
… Arithmetic codeword to be encrypted, 113… Common key for encryption, 1
14 ... Encrypted arithmetic code word, 115 ... Encrypted non-target arithmetic code word, 116 ... Encrypted target information, 117 ... Packetized coded code stream, 118 ... Encrypted target arithmetic code word, 119 ... Encrypted non-target arithmetic codeword, 120 ... Encrypted arithmetic codeword, 121 ... Packetized coded code stream, 122 ... Packet separated into encryption target and non-target, 123 ... Bit Rate information, 124 ... packet header before bit rate conversion, 125 ... packet header after rewriting, 126 ... packet body after bit rate conversion, 127 ... packet after bit rate conversion, 128 ... encryption target and non-target Separated packets, 129 ... Arithmetically decoded quantized coefficients, 130 ... Bit modeling decoded quantized coefficients, 131 ... Code Restored quantized coefficients from the lock to the image plane, 132 ... wavelet transform coefficients after inverse quantization, 133 ... decoded image after the inverse wavelet transform.

   ─────────────────────────────────────────────────── ─── Continued front page    (72) Inventor Seiji Kimura             6-735 Kita-Shinagawa, Shinagawa-ku, Tokyo Soni             -Inside the corporation (72) Inventor Hitoshi Kiya             5-9-3-307 Minami Osawa, Hachioji City, Tokyo F term (reference) 5C059 KK43 MA00 MA24 MA35 MC11                       MC38 ME11 PP01 RB02 RB17                       RC07 RC11 RC24 RC28 RC35                       SS06 SS12 TA00 TA43 TB00                       TC04 TC18 TC47 UA02 UA05                       UA15 UA32                 5C064 CA14 CC04                 5C078 AA04 BA53 CA02 CA47 DA01                       DB16 DB19                 5J064 AA02 BA15 BC08 BC16 BD02                 5J104 AA12 AA15 JA03 NA14 PA07

Claims (22)

[Claims] 1. Encoding means for encoding an input image signal according to a predetermined image processing standard, and means for separating / extracting an encoded code stream to be encrypted from the encoded code stream. An image encoding apparatus comprising: an encryption unit that performs an encryption process on the extracted encoded code stream; and a unit that writes the encrypted data back to the same portion of the original encoded code stream. 2. The image coding apparatus according to claim 1, wherein the coded code stream to be encrypted is an arithmetic code which is output data of EBCOT defined by a predetermined image processing standard. 3. The image coding apparatus according to claim 1, wherein the coded code stream to be encrypted is coded data of a packet body defined by a predetermined image processing standard. 4. The image coding apparatus according to claim 3, wherein the coded code stream to be encrypted is coded data of a predetermined code block in the packet body. 5. The image coding apparatus according to claim 4, wherein the coded code stream to be encrypted is coded data of a predetermined coding pass in the code block. 6. The image coding apparatus according to claim 1, wherein said encryption means is block encryption means using common key encryption. 7. The image coding apparatus according to claim 6, wherein the cipher block size of said block cipher means is 1 byte, and 1 byte of encrypted data is output for 1 byte of input data. 8. The means for separating / extracting the coded codestream includes a means for reading a Scod parameter defined in the JPEG-2000 standard, a means for decoding a packet header of the codestream, and a means for decoding the codestream. The image coding apparatus according to claim 1, further comprising means for detecting a packet body. 9. The packet body detecting means decodes the packet header when it is determined that the SOP marker and the EPH marker do not exist in the encoded code stream as a result of reading the Scod parameter. 9. The image coding apparatus according to claim 8, further comprising means for detecting a packet body existing next to the packet header from the information on the packet header length. 10. The means for separating / extracting the coded code stream comprises means for reading out Scod parameters defined by the JPEG-2000 standard, and SGcod, Csiz, SPcod defined by the JPEG-2000 standard. 3. The method according to claim 1, further comprising: a means for detecting each parameter of step 1, a means for detecting a sequence number from the Nsop parameter in the SOP, and a means for detecting a packet body.
The image encoding device described. 11. The packet means detecting means detects the Scod parameter and, as a result of reading the Scod parameter, determines from the Nsop parameter in the SOP if it is determined that an SOP marker and an EPH marker are present in the encoded code stream. The image coding apparatus according to claim 10, further comprising means for detecting a packet body to be encrypted from the detected sequence number. 12. The image coding apparatus according to claim 1, wherein the coded code stream to be encrypted is coded data of a code block corresponding to a predetermined image area. 13. The image coding apparatus according to claim 1, wherein the coded code stream to be encrypted is coded data of a predetermined coding pass in a code block corresponding to a predetermined image area. 14. The image coding apparatus according to claim 1, wherein the coded code stream to be encrypted is coded data of a packet body existing at a predetermined resolution level. 15. The image coding according to claim 1, wherein the coded code stream to be encrypted is coded data of a predetermined code block among coded data of a packet body existing at a predetermined resolution level. apparatus. 16. The image coding according to claim 1, wherein the coded code stream to be encrypted is coded data of a predetermined coding pass among code blocks in a packet body existing at a predetermined resolution level. apparatus. 17. The image coding apparatus according to claim 1, wherein the coded code stream to be encrypted is coded data of a packet body existing in a predetermined layer. 18. The image coding apparatus according to claim 1, wherein the coded code stream to be encrypted is coded data of a code block of a packet body existing in a predetermined layer. 19. The image coding apparatus according to claim 1, wherein the coded code stream to be encrypted is coded data of a predetermined coding path among code blocks in a packet body existing in a predetermined layer. . 20. Transform means for wavelet transforming an input image signal, quantizing means for quantizing the transform result of the wavelet transforming means, and quantized coefficients obtained by the quantizing means from the most significant bit to the least significant bit. A means for expanding the bit plane to reach a bit, a means for performing arithmetic coding processing on the bit plane for each predetermined entropy coding processing unit, and a predetermined arithmetic code for the arithmetic code generated by the arithmetic coding processing. With a bit rate control means, a means for creating a header in the arithmetic code whose bit rate is controlled, a means for encrypting the arithmetic code, and an arithmetic code after the encryption and the header. An image coding apparatus having a packet generation means for generating a packet. 21. A means for inputting a partially-encrypted coded code stream to separate / extract a partially-encoded coded stream; and the extracted partially-encrypted coded data, Decryption means for deciphering using a common key, means for arithmetically decoding the unencrypted encoded data in the encoded code stream, and the encoded data decrypted by the decryption means. An image decoding apparatus having: 22. A packet separating means for separating a packet header and a packet body from an input partially encrypted code stream, and truncating a part of encoded data of a code block or a coding path in the packet body. An image having means, means for rewriting the code block and coding path information of the coded code stream after the truncation processing in a packet header, and means for generating a new packet from the rewritten packet header and packet body Decoding device.