Privacy preserving method based on compressive sensing stepwise prediction

Abstract

The application provides a privacy protection method based on compressed sensing step-by-step prediction, which comprises the following steps: constructing a deterministic scalar measurement matrix according to the size of an original image; obtaining reconstruction information of the image in a compressed domain based on the deterministic scalar measurement matrix; obtaining a prediction value by using a pre-reconstruction decoder and using the prediction value to obtain global information of the original image; using an edge predictor to obtain a final prediction error pixel set in combination with a preliminary error pixel set; performing compressed bit stream processing on the final prediction error pixel set to obtain a compressed image; encrypting the compressed image to generate a marked encrypted image; and decrypting the marked encrypted image by using an encryption key at a receiving end to obtain a lossless image and secret information. By constructing an edge predictor to calculate a final prediction error pixel set, the application can greatly remove redundant information and improve the embedding capacity of secret information.

Description

Technical Field

[0001] This invention relates to the fields of image processing and information security technology, and in particular to a privacy protection method based on progressive prediction using compressed sensing. Background Technology

[0002] Advances in internet technology have fundamentally changed human behavior. With the increasing prevalence of tablets and smartphones, users frequently store their data on third-party cloud storage platforms such as Google Drive, Dropbox, and OneDrive for convenient and quick access to the data they need.

[0003] However, in this highly digital world, the existence of spoofed communications has become a serious problem due to the rapid emergence of unauthorized data access and violations of network policies. These issues have led to the use of encryption technology in applications as a means of protecting data. However, because encrypted data can be detected by sophisticated intruders during transmission, this can raise suspicion and lead to unauthorized access. Summary of the Invention

[0004] In view of the above, the main objective of this invention is to propose a privacy protection method based on compressed sensing progressive prediction in order to solve the aforementioned technical problems.

[0005] This invention proposes a privacy protection method based on compressed sensing progressive prediction, the method comprising the following steps: Step 1: Obtain the original image and construct a deterministic scalar measurement matrix based on the dimensions of the original image; The signal of the original image is compressed using 2D compressed sensing and a deterministic scalar measurement matrix to obtain the image reconstruction information in the compressed domain; Step 2: Reconstruct the image information in the compressed domain using a pre-reconstruction decoder and obtain the predicted value. Use the predicted value to recover the global information of the original image. A preliminary error pixel set is obtained from the original image and the predicted value. Then, the final prediction error pixel set is obtained by combining the preliminary error pixel set with the edge predictor. Step 3: Divide the final prediction error pixel set into blocks, process the result using a bit-plane rearrangement compression algorithm, and then compress the result using a compressed bit stream to obtain a compressed image. Step 4: Generate a pseudo-random sequence using the encryption key, and generate an encrypted image using the compressed image and the pseudo-random sequence; A pseudo-random sequence is generated using an encryption key, and encrypted password information is generated using secret information and the pseudo-random sequence. A tagged encrypted image is generated by combining the encrypted image and the encrypted password information; Step 5: The receiving end decrypts the marked encrypted image using the corresponding encryption key to obtain the final prediction error pixel set. After inputting the final prediction error pixel set into the edge predictor, the reconstruction information of the image in the compressed domain and the preliminary error pixel set are obtained. The lossless image and secret information are obtained by using the reconstruction information of the image in the compressed domain, the preliminary error pixel set and the global information of the original image.

[0006] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. This invention constructs an edge predictor to calculate its final prediction error pixel set, thereby greatly removing redundant information and improving the embedding capacity of secret information; 2. This invention not only ensures privacy and has independence and separability in data extraction and image restoration, but also achieves low sampling complexity and saves resource storage costs, reducing the burden on equipment.

[0007] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by means of embodiments of the invention. Attached Figure Description

[0008] Figure 1 This is a flowchart of the privacy protection method based on compressed sensing and progressive prediction proposed in this invention.

[0009] Figure 2 This is a schematic diagram of the framework of the privacy protection method based on compressed sensing and stepwise prediction proposed in this invention. Detailed Implementation

[0010] Embodiments of the present invention are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0011] These and other aspects of the embodiments of the present invention will become clear from the following description and accompanying drawings. In these descriptions and drawings, some specific embodiments of the present invention are specifically disclosed to illustrate some ways of implementing the principles of the embodiments of the present invention; however, it should be understood that the scope of the embodiments of the present invention is not limited thereto.

[0012] Please see Figure 1 and Figure 2 This invention proposes a privacy protection method based on compressed sensing progressive prediction, which includes the following steps: Step 1: Obtain the original image and construct a deterministic scalar measurement matrix based on the dimensions of the original image; The signal of the original image is compressed using 2D compressed sensing and a deterministic scalar measurement matrix to obtain the image reconstruction information in the compressed domain; In step 1, the original image is acquired, and a deterministic scalar measurement matrix is ​​constructed based on the dimensions of the original image. The corresponding relationship in the process is: ; in, Represents a deterministic scalar measurement matrix. Represents non-zero element values; The signal of the original image is compressed using 2D compressed sensing and a deterministic scalar measurement matrix to obtain the reconstructed information of the image in the compressed domain. The corresponding relationship in this process is as follows: ; in, This represents the reconstructed information of the image in the compressed domain. and Both represent deterministic scalar measurement matrices. This represents the original image.

[0013] Step 2: Reconstruct the image information in the compressed domain using a pre-reconstruction decoder and obtain the predicted value. Use the predicted value to recover the global information of the original image. A preliminary error pixel set is obtained from the original image and the predicted value. Then, the final prediction error pixel set is obtained by combining the preliminary error pixel set with the edge predictor. In step 2, the reconstruction information of the image in the compressed domain is reconstructed using a pre-reconstruction decoder to obtain predicted values. These predicted values ​​are then used to recover the global information of the original image. The expression for the pre-reconstruction decoder is: ; in, Indicates the predicted value. This represents a 2D compressed sensing reconstruction algorithm. The bit depth of the predicted value. This indicates taking the maximum value. This indicates taking the minimum value. This indicates rounding to the nearest integer. A preliminary error pixel set is obtained from the original image and the predicted values. Then, using an edge predictor, the final prediction error pixel set is obtained by combining the preliminary error pixel set. The expression for the edge predictor is: ; in, Indicates the predicted value. Represents the first of the matrix OK, Represents the first of the matrix List, Represents the original pixel. This represents the pixel value located at the top left corner of the original pixel. This represents the pixel value located above the original pixel. This represents the pixel value to the left of the original pixel; A preliminary error pixel set is obtained from the original image and the predicted values. Then, using an edge predictor, the final prediction error pixel set is obtained by combining the preliminary error pixel set. The corresponding relationship in this process is as follows: ; in, Indicates the pixel in the final prediction error set. Line 1 The corresponding column The pixel value of the bit. This indicates a round-down operation. This indicates a modulo operation.

[0014] Step 3: Divide the final prediction error pixel set into blocks, process the result using a bit-plane rearrangement compression algorithm, and then compress the result using a compressed bit stream to obtain a compressed image. In step 3, the final prediction error pixel set is divided into blocks, and the processing result is processed by a bit-plane rearrangement compression algorithm. Then, the processing result is compressed using a compressed bitstream to obtain a compressed image. The relationship of the bit-plane rearrangement compression algorithm is as follows: ; in, , , and These represent the bit streams extracted after four different bit stream rearrangement methods. Bit, This represents the bit-plane matrix after the final prediction error pixel set has been processed by dividing it into blocks.

[0015] Step 4: Generate a pseudo-random sequence using the encryption key, and generate an encrypted image using the compressed image and the pseudo-random sequence; A pseudo-random sequence is generated using an encryption key, and encrypted password information is generated using secret information and the pseudo-random sequence. A tagged encrypted image is generated by combining the encrypted image and the encrypted password information; In step 4, using the encryption key Generate pseudo-random sequences The encrypted image is generated by compressing the image and using a pseudo-random sequence. The corresponding relationship in the process is as follows: ; in, Indicates an encrypted image. Indicates a compressed image. Represents a pseudo-random sequence. This represents the XOR operation; By encryption key Generate pseudo-random sequences The process of generating encrypted cipher information using secret information and pseudo-random sequences has the following relationship: ; in, This indicates encrypted password information. Indicates secret information, Represents a pseudo-random sequence. This represents the XOR operation.

[0016] Step 5: The receiving end decrypts the marked encrypted image using the corresponding encryption key to obtain the final prediction error pixel set. After inputting the final prediction error pixel set into the edge predictor, the reconstruction information of the image in the compressed domain and the preliminary error pixel set are obtained. The lossless image and secret information are obtained by using the reconstruction information of the image in the compressed domain, the preliminary error pixel set and the global information of the original image.

[0017] Furthermore, in step 5, the lossless image and secret information can be obtained, depending on the number of keys at the receiving end. Specifically, there are three cases: Scenario 1: The recipient only has the encryption key Because of the encryption key The process of embedding secret information is not affected; only the image is encrypted. Therefore, the extracted secret information can be encrypted using the encryption key. Successfully decrypted, no encryption key available. The original image content cannot be recovered; Scenario 2: The recipient only has the encryption key. The original image can be recovered. First, the block size and the number and location of overflow pixels are extracted from the most significant bit plane. It is then determined whether the bitstream has been compressed; the bitstream is decoded to obtain the uncompressed bitstream. Next, the number of compressed bitstreams in each bit plane is extracted from the least significant bit plane. Finally, the image is processed using the encryption key. Decrypt using the decryption algorithm. No encryption key is available. Therefore, it is impossible to correctly extract secret information; Scenario 3: The recipient possesses the encryption key. and It can be done through encryption keys. The embedded secret information can be correctly extracted and can be obtained through the encryption key. Restore the original image.

[0018] Furthermore, according to Table 1, the average PSNR and SSIM of different original images can be seen.

[0019] Table 1. PSNR and SSIMS of reconstructed image quality

[0020] Meanwhile, according to Table 2, the embedding capacity of different 512×512 original images is as follows: Table 2. Embedding capacity (bits) of this model

[0021] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0022] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0023] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A privacy protection method based on compressed sensing and progressive prediction, characterized in that, The method includes the following steps: Step 1: Obtain the original image and construct a deterministic scalar measurement matrix based on the dimensions of the original image; The signal of the original image is compressed using 2D compressed sensing and a deterministic scalar measurement matrix to obtain the image reconstruction information in the compressed domain; Step 2: Reconstruct the image information in the compressed domain using a pre-reconstruction decoder and obtain the predicted value. Use the predicted value to recover the global information of the original image. A preliminary error pixel set is obtained from the original image and the predicted value. Then, the final prediction error pixel set is obtained by combining the preliminary error pixel set with the edge predictor. Step 3: Divide the final prediction error pixel set into blocks, process the result using a bit-plane rearrangement compression algorithm, and then compress the result using a compressed bit stream to obtain a compressed image. Step 4: Generate a pseudo-random sequence using the encryption key, and generate an encrypted image using the compressed image and the pseudo-random sequence; A pseudo-random sequence is generated using an encryption key, and encrypted password information is generated using secret information and the pseudo-random sequence. A tagged encrypted image is generated by combining the encrypted image and the encrypted password information; Step 5: The receiving end decrypts the marked encrypted image using the corresponding encryption key to obtain the final prediction error pixel set. After inputting the final prediction error pixel set into the edge predictor, the reconstruction information of the image in the compressed domain and the preliminary error pixel set are obtained. The lossless image and secret information are obtained by using the reconstruction information of the image in the compressed domain, the preliminary error pixel set and the global information of the original image.

2. The privacy protection method based on compressed sensing progressive prediction according to claim 1, characterized in that, In step 1, the original image is acquired, and a deterministic scalar measurement matrix is ​​constructed based on the dimensions of the original image. The corresponding relationship in the process is: ； in, Represents a deterministic scalar measurement matrix. This represents a non-zero element value.

3. The privacy protection method based on progressive prediction using compressed sensing according to claim 2, characterized in that, In step 1, the signal of the original image is compressed using 2D compressed sensing and a deterministic scalar measurement matrix to obtain the reconstructed information of the image in the compressed domain. The corresponding relationship in this process is as follows: ； in, This represents the reconstructed information of the image in the compressed domain. and Both represent deterministic scalar measurement matrices. This represents the original image.

4. The privacy protection method based on progressive prediction using compressed sensing according to claim 3, characterized in that, In step 2, the reconstruction information of the image in the compressed domain is reconstructed using a pre-reconstruction decoder to obtain predicted values. These predicted values ​​are then used to recover the global information of the original image. The expression for the pre-reconstruction decoder is: ； in, Indicates the predicted value. This represents a 2D compressed sensing reconstruction algorithm. Table 4 shows the bit depth of the predicted values. This indicates taking the maximum value. This indicates taking the minimum value. This means rounding to the nearest integer.

5. The privacy protection method based on compressed sensing progressive prediction according to claim 4, characterized in that, In step 2, a preliminary error pixel set is obtained from the original image and the predicted values. Then, an edge predictor is used to combine this preliminary error pixel set with the final prediction error pixel set. The expression for the edge predictor is: ； in, Indicates the predicted value. Represents the first of the matrix OK, Represents the first of the matrix List, Represents the original pixel. This represents the pixel value located at the top left corner of the original pixel. This represents the pixel value located above the original pixel. This represents the pixel value located to the left of the original pixel.

6. The privacy protection method based on progressive prediction using compressed sensing according to claim 5, characterized in that, In step 2, a preliminary error pixel set is obtained from the original image and the predicted value. Then, using an edge predictor, the final prediction error pixel set is obtained by combining the preliminary error pixel set. The corresponding relationship in this process is: ； in, Indicates the pixel in the final prediction error set. Line number The corresponding column The pixel value of the bit. This indicates a round-down operation. This indicates a modulo operation.

7. The privacy protection method based on progressive prediction using compressed sensing according to claim 6, characterized in that, In step 3, the final prediction error pixel set is divided into blocks, the processing result is processed by a bit plane rearrangement compression algorithm, and then the processing result is compressed by a compressed bit stream to obtain a compressed image. The relationship for the bit-plane rearrangement compression algorithm is as follows: ； in, , , and These represent the bit streams extracted after four different bit stream rearrangement methods. Bit, This represents the bit-plane matrix after the final prediction error pixel set has been processed by dividing it into blocks.

8. The privacy protection method based on progressive prediction using compressed sensing according to claim 7, characterized in that, In step 4, a pseudo-random sequence is generated using the encryption key, and an encrypted image is generated using the compressed image and the pseudo-random sequence. The corresponding relationship in this process is as follows: ； in, Indicates an encrypted image. Indicates a compressed image. Represents a pseudo-random sequence. This represents the XOR operation.

9. The privacy protection method based on progressive prediction using compressed sensing according to claim 8, characterized in that, In step 4, a pseudo-random sequence is generated using the encryption key, and encrypted cipher information is generated using the secret information and the pseudo-random sequence. The corresponding relationship in this process is as follows: ； in, This indicates encrypted password information. Indicates secret information, Represents a pseudo-random sequence. This represents the XOR operation.

Images