Image data acquisition method and system for an image processing chip

Abstract

The present invention discloses an image data acquisition method and system for an image processing chip, which relates to the field of images and solves the problems of poor transmission security and poor display effect existing in the image data acquisition method. The method includes: Step S1: respectively acquire multiple terminal images, respectively perform privacy analysis on each terminal image, and obtain the image privacy area corresponding to each terminal image to obtain image privacy data; Step S2: perform image quality analysis on each terminal image according to the image privacy data, and obtain multiple image quality coefficients to obtain image quality data; Step S3: perform encrypted transmission and image quality optimization on each terminal image according to the image privacy data and the image quality data. The present invention can improve the image transmission security and the image display quality.

Description

Technical Field

[0001] The present invention belongs to the field of images and relates to chip technology, specifically an image data acquisition method and system for an image processing chip. Background Art

[0002] Existing image data acquisition methods have the following defects when performing image acquisition and transmission:

[0003] 1. When existing image acquisition methods perform image acquisition and transmission, they cannot identify and encrypt privacy areas in the image, resulting in certain security risks during transmission;

[0004] 2. When existing image acquisition methods perform image output, they cannot optimize the image quality according to the performance parameters of the image output device, resulting in poor image display effects;

[0005] Therefore, we propose an image data acquisition method and system for an image processing chip. Summary of the Invention

[0006] Aiming at the deficiencies of the existing technology, the purpose of the present invention is to provide an image data acquisition method and system for an image processing chip. The present invention is based on separately acquiring and collecting multiple terminal images, separately performing privacy analysis on each terminal image, and obtaining the image privacy area corresponding to each terminal image to obtain image privacy data. According to the image privacy data, image quality analysis is performed on each terminal image, and multiple image quality coefficients are obtained to obtain image quality data. According to the image privacy data and image quality data, each terminal image is processed into an encrypted terminal image to obtain multiple encrypted terminal images, and the encrypted terminal images are transmitted, and the image quality of each terminal image is optimized according to the image quality data;

[0007] To achieve the above purpose, the present invention adopts the following technical solutions: An image data acquisition method for an image processing chip specifically includes the following steps:

[0008] Step S1: Separately acquire and collect multiple terminal images, separately perform privacy analysis on each terminal image, and obtain the image privacy area corresponding to each terminal image to obtain image privacy data;

[0009] Step S2: Perform image quality analysis on each terminal image according to the image privacy data, and obtain multiple image quality coefficients to obtain image quality data;

[0010] Step S3: Process each terminal image into an encrypted terminal image according to the image privacy data and image quality data to obtain multiple encrypted terminal images, transmit the encrypted terminal images, and optimize the image quality of each terminal image according to the image quality data.

[0011] Further, in step S1, the following specific steps are further included:

[0012] Step S11: Obtain the images stored in the image processing terminal where the image processing chip is located, and name the stored images as the first terminal image to the a-th terminal image according to the saving time;

[0013] Step S12: Mark the privacy areas of the first terminal image to the a-th terminal image to obtain the first image privacy data to the a-th image privacy area;

[0014] Step S13: Randomly select one terminal image from the first terminal image to the a-th terminal image as the sample privacy image;

[0015] Step S14: Perform privacy marking on the sample privacy image to obtain the image privacy area;

[0016] Step S15: Respectively obtain the image privacy areas of the first terminal image to the a-th terminal image to obtain the first image privacy area to the a-th image privacy area;

[0017] Step S16: Define the first image privacy area to the a-th image privacy area as the image privacy data.

[0018] Further, in step S14, the following specific steps are further included:

[0019] Step S141: Perform layer scanning on the sample privacy image through OCR technology, and divide the image content in the sample privacy image into the first image content and the second image content;

[0020] Step S142: Respectively set multiple privacy keywords, use text recognition algorithms to perform text analysis on the first image content, mark the area in the first image content with the same text as any privacy keyword as the image privacy area, and mark the image privacy area;

[0021] Step S143: Perform privacy content marking on the second image content to obtain the image privacy area corresponding to the second image content.

[0022] Further, in step S143, the following specific steps are further included:

[0023] Step S1431: Use web crawler technology to obtain multiple images with privacy content as sample privacy image data, and perform annotation on the privacy area corresponding to each image of the sample privacy image data through manual annotation to obtain the sample privacy image annotation data;

[0024] Step S1432: Divide multiple images in the sample privacy image data into an image training set and an image test set according to the image division ratio;

[0025] Step S1433: Create an image recognition model through an existing artificial intelligence platform, and use the image training set to perform image recognition training on the image recognition model until the image recognition model completes image recognition training for each image in the image training set;

[0026] Step S1434: Use the image test set to perform training tests on the image recognition model, obtain the recognition accuracy, and get the current recognition accuracy. Obtain the benchmark recognition accuracy. If the current recognition accuracy is greater than or equal to the benchmark recognition accuracy, the training of the image recognition model is completed to obtain a privacy content recognition model. If the current recognition accuracy is less than the benchmark recognition accuracy, use the image training set to continue training the image recognition model until the current recognition accuracy is greater than or equal to the benchmark recognition accuracy;

[0027] Step S1435: Use the privacy content recognition model to perform privacy recognition on the second image content, and mark the recognized privacy area as the image privacy area.

[0028] Furthermore, in the said Step S2, the following specific steps are further included:

[0029] Step S21: Obtain image privacy data, and respectively obtain the first terminal image to the a-terminal image according to the image privacy data;

[0030] Step S22: Randomly select an image from the first terminal image to the a-terminal image as the sample image quality analysis image;

[0031] Step S23: Perform image quality analysis on the sample image quality analysis image to obtain the image quality coefficient corresponding to the sample image quality analysis image;

[0032] Step S24: Perform image quality analysis on the first terminal image to the a-terminal image respectively to obtain the image quality coefficients corresponding to the first terminal image to the a-terminal image respectively;

[0033] Step S25: Define the image quality coefficients corresponding to the first terminal image to the a-terminal image respectively as image quality data.

[0034] Furthermore, in the said Step S23, the following specific steps are further included:

[0035] Step S231: Obtain the original resolution value corresponding to the sample image quality analysis image;

[0036] Step S232: Obtain the original image contrast corresponding to the sample image quality analysis image;

[0037] Step S233: Use an edge detection algorithm to perform edge detection on the sample image quality analysis image to obtain the edge intensity value corresponding to the sample image quality analysis image;

[0038] Step S234: Calculate the image quality coefficient corresponding to the sample image quality analysis image through the edge intensity value, the original image contrast, and the original resolution value;

[0039] Calculate the image quality coefficient corresponding to the sample image quality analysis image. The specific formula is as follows:

[0040] ;

[0041] where Hzx is the image quality coefficient, Byq is the edge intensity value, Dbd is the original image contrast, and Fbl is the original resolution value;

[0042] The step S231 further includes the following specific steps:

[0043] Step S2311: Randomly select multiple pixel distribution rows in the sample image quality analysis image, respectively obtain the pixel point values corresponding to each pixel distribution row to get multiple horizontal pixel point quantity values, and calculate the average of the obtained multiple horizontal pixel point quantity values to obtain the average horizontal pixel point quantity value;

[0044] Step S2312: Randomly select multiple pixel distribution columns in the sample image quality analysis image, respectively obtain the pixel point values corresponding to each pixel distribution column to get multiple vertical pixel point quantity values, and calculate the average of the obtained multiple vertical pixel point quantity values to obtain the average vertical pixel point quantity value;

[0045] Step S2313: Calculate the product of the average horizontal pixel point quantity value and the average vertical pixel point quantity value to obtain the original resolution value corresponding to the sample image quality analysis image.

[0046] Further, in the step S232, it further includes the following specific steps:

[0047] Step S2321: Randomly select m monitoring pixel points in the sample image quality analysis image, respectively obtain the brightness values corresponding to each pixel point to get multiple pixel point brightness values;

[0048] Step S2322: Compare the magnitudes of the obtained multiple pixel point brightness values, mark the pixel point brightness value with the largest value as the first characteristic pixel point brightness value, and mark the pixel point brightness value with the smallest value as the second characteristic pixel point brightness value;

[0049] Step S2323: Calculate the original image contrast corresponding to the sample image quality analysis image from the brightness values of the first characteristic pixel points and the second characteristic pixel points;

[0050] Calculate the original image contrast corresponding to the sample image quality analysis image, and the specific formula configuration is as follows:

[0051] ;

[0052] Where Dbd is the original image contrast corresponding to the sample image quality analysis image, Ld1 is the brightness value of the first characteristic pixel point, and Ld2 is the brightness value of the second characteristic pixel point.

[0053] Furthermore, in step S3, the following specific steps are further included:

[0054] Step S31: Obtain image privacy data, and respectively obtain the first image privacy area to the a-th image privacy area according to the image privacy data;

[0055] Step S32: Perform privacy encryption on the first image privacy area to the a-th image privacy area to obtain the first image privacy encrypted area to the a-th image privacy encrypted area, and respectively cover the first image privacy area to the a-th image privacy area through the first image privacy encrypted area to the a-th image privacy encrypted area to obtain the first encrypted terminal image to the a-th encrypted terminal image, and transmit the first encrypted terminal image to the a-th encrypted terminal image to the receiving end user;

[0056] Step S33: Obtain image quality data, and respectively obtain the image quality coefficients corresponding to the first terminal image to the a-th terminal image according to the image quality data;

[0057] Step S34: Perform image quality adjustment on the first terminal image;

[0058] In step S34, the following specific steps are further included:

[0059] Step S341: Respectively obtain the calibrated screen contrast, calibrated screen edge intensity value, and calibrated screen resolution value of the image display device corresponding to the first terminal image;

[0060] Step S342: Calculate the screen display quality coefficient corresponding to the image display device from the calibrated screen contrast, calibrated screen edge intensity value, and calibrated screen resolution value;

[0061] Calculate the screen display quality coefficient, and the specific formula is as follows:

[0062] ;

[0063] Among them, Pzx is the screen display quality coefficient, Pyq is the calibrated screen edge intensity value, Pbd is the calibrated screen contrast, and Pbl is the calibrated screen resolution value;

[0064] Step S343: Obtain the image quality coefficient corresponding to the first terminal image;

[0065] Step S344: When the image quality coefficient is greater than or equal to the screen display quality coefficient, directly output the first terminal image through the display device;

[0066] Step S345: When the image quality coefficient is less than the screen display quality coefficient, directly optimize the display quality of the first terminal image until the image quality coefficient is equal to the screen display quality coefficient;

[0067] Step S345: Adjust the display quality of the second terminal image to the a-th terminal image respectively.

[0068] Furthermore, in the step S32, the following specific steps are further included:

[0069] Step S321: Use a random number generator to generate a byte sequence with a length of 128 bits as the generation key and send the generation key to the data decryption unit;

[0070] Step S322: Convert the first image privacy area into the first image privacy area Unicode encoding through the Unicode decoder, and then convert the first image privacy area Unicode encoding into the first image privacy area binary encoding. Divide the first image privacy area binary encoding into data groups with equal lengths. Each data group has a length of 128 bits, corresponding to the 128-bit byte sequence key;

[0071] Step S323: Use the previous data group of the current data group as the initial vector, perform an exclusive OR operation (different is one, the same is zero) on the current data group and the initial vector, and then encrypt the result of the exclusive OR operation with the generation key. Use the encrypted data as the initial vector of the next data group, repeat the above process, and use the last encrypted data as the initial vector of the first data group to complete the encryption of all data groups, realize the encryption of the first image privacy area, and obtain the first image privacy encryption area;

[0072] Step S324: Encrypt the second image privacy area to the a-th image privacy area respectively to obtain the second image privacy encryption area to the a-th image privacy encryption area.

[0073] An image data acquisition system for an image processing chip, and the specific working process of each module is as follows:

[0074] Privacy Data Module: It is used to separately acquire and collect multiple terminal images, perform privacy analysis on each terminal image respectively, obtain the image privacy area corresponding to each terminal image, and obtain image privacy data;

[0075] Image Quality Module: It is used to perform image quality analysis on each terminal image according to the image privacy data, and obtain multiple image quality coefficients to obtain image quality data;

[0076] Data Processing Module: It is used to process each terminal image into an encrypted terminal image according to the image privacy data and the image quality data, obtain multiple encrypted terminal images, transmit the encrypted terminal images, and optimize the image quality of each terminal image according to the image quality data.

[0077] In summary, due to the adoption of the above technical solutions, the beneficial effects of the present invention are as follows:

[0078] 1. By automatically marking and encrypting the privacy area in the terminal image and using the encrypted terminal image for transmission, the present invention can effectively improve the security of image data transmission;

[0079] 2. When outputting the image through the image output device, by obtaining the screen display quality coefficient corresponding to the output device and comparing it with the image quality coefficient, and optimizing the image according to the comparison result, the present invention can ensure the output quality of the image. BRIEF DESCRIPTION OF THE DRAWINGS

[0080] For the convenience of those skilled in the art to understand, the present invention will be further described below with reference to the accompanying drawings.

[0081] Figure 1 It is the implementation step diagram of the present invention;

[0082] Figure 2 It is the overall system block diagram of the present invention;

[0083] Figure 3 It is the schematic diagram of the sample privacy image in the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0084] The technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0085] Embodiment 1

[0086] Please refer to Figure 1, the present invention provides a technical solution: an image data acquisition method and system for an image processing chip, including the following specific steps:

[0087] Step S1: Obtain multiple terminal images respectively, perform privacy analysis on each terminal image respectively, and obtain the image privacy area corresponding to each terminal image to obtain image privacy data;

[0088] Step S11: Obtain the images stored in the image processing terminal where the image processing chip is located, and name the stored images as the first terminal image to the a-th terminal image according to the saving time respectively;

[0089] Step S12: Mark the privacy areas of the first terminal image to the a-th terminal image to obtain the first image privacy data to the a-th image privacy area;

[0090] Step S13: Randomly select a terminal image from the first terminal image to the a-th terminal image as a sample privacy image;

[0091] Step S14: Perform privacy marking on the sample privacy image to obtain an image privacy area;

[0092] In the said step S14, it further includes the following specific steps:

[0093] Step S141: Perform layer scanning on the sample privacy image through OCR technology, and divide the image content in the sample privacy image into the first image content and the second image content;

[0094] Step S142: Set multiple privacy keywords respectively, use text recognition algorithms to perform text analysis on the first image content, mark the areas in the first image content that are the same text as any privacy keyword as the image privacy area, and mark the image privacy area;

[0095] Step S143: Perform privacy content marking on the second image content;

[0096] In the said step S143, it further includes the following specific steps:

[0097] Step S1431: Obtain multiple images with privacy content as sample privacy image data through web crawler technology, and annotate the privacy areas corresponding to each image in the sample privacy image data through manual annotation to obtain sample privacy image annotation data;

[0098] Step S1432: Divide the multiple images in the sample privacy image data into an image training set and an image test set according to the image division ratio;

[0099] Step S1433: Create an image recognition model through an existing artificial intelligence platform, and use the image training set to perform image recognition training on the image recognition model until the image recognition model has completed image recognition training for each image in the image training set;

[0100] Step S1434: Use the image test set to conduct training tests on the image recognition model, obtain the recognition accuracy, and get the current recognition accuracy. Obtain the benchmark recognition accuracy. If the current recognition accuracy is greater than or equal to the benchmark recognition accuracy, the training of the image recognition model is completed, and a privacy content recognition model is obtained. If the current recognition accuracy is less than the benchmark recognition accuracy, use the image training set to continue training the image recognition model until the current recognition accuracy is greater than or equal to the benchmark recognition accuracy;

[0101] Step S1435: Use the privacy content recognition model to perform privacy recognition on the second image content, and mark the recognized privacy area as the image privacy area;

[0102] Step S15: Obtain the image privacy areas for the first terminal image to the a-th terminal image respectively, and get the first image privacy area to the a-th image privacy area;

[0103] Step S16: Define the first image privacy area to the a-th image privacy area as the image privacy data;

[0104] Step S2: Perform image quality analysis on each terminal image according to the image privacy data, and obtain multiple image quality coefficients to get the image quality data;

[0105] Step S21: Obtain the image privacy data, and respectively obtain the first terminal image to the a-th terminal image according to the image privacy data;

[0106] Step S22: Randomly select an image from the first terminal image to the a-th terminal image as the sample image for quality analysis;

[0107] Step S23: Perform quality analysis on the sample image for quality analysis to obtain the image quality coefficient corresponding to the sample image for quality analysis;

[0108] The step S23 further includes the following specific steps:

[0109] Step S231: Obtain the original resolution value corresponding to the sample image for quality analysis;

[0110] The step S231 further includes the following specific steps:

[0111] Step S2311: Randomly select multiple pixel distribution rows in the sample image quality analysis image, respectively obtain the pixel point values corresponding to each pixel distribution row, obtain multiple horizontal pixel point quantity values, and calculate the average value of the obtained multiple horizontal pixel point quantity values to obtain the average horizontal pixel point quantity value;

[0112] Step S2312: Randomly select multiple pixel distribution columns in the sample image quality analysis image, respectively obtain the pixel point values corresponding to each pixel distribution column, obtain multiple vertical pixel point quantity values, and calculate the average value of the obtained multiple vertical pixel point quantity values to obtain the average vertical pixel point quantity value;

[0113] Step S2313: Calculate the product of the average horizontal pixel point quantity value and the average vertical pixel point quantity value to obtain the original resolution value corresponding to the sample image quality analysis image;

[0114] Step S232: Obtain the original image contrast corresponding to the sample image quality analysis image;

[0115] In the said step S232, the following specific steps are further included:

[0116] Step S2321: Randomly select m monitoring pixel points in the sample image quality analysis image, respectively obtain the brightness values corresponding to each pixel point to obtain multiple pixel point brightness values;

[0117] Step S2322: Compare the magnitudes of the obtained multiple pixel point brightness values, mark the pixel point brightness value with the largest value as the first characteristic pixel point brightness value, and mark the pixel point brightness value with the smallest value as the second characteristic pixel point brightness value;

[0118] Step S2323: Calculate the original image contrast corresponding to the sample image quality analysis image through the first characteristic pixel point brightness value and the second characteristic pixel point brightness value;

[0119] Calculate the original image contrast corresponding to the sample image quality analysis image, and the specific formula configuration is as follows:

[0120] ;

[0121] Wherein, Dbd is the original image contrast corresponding to the sample image quality analysis image, Ld1 is the first characteristic pixel point brightness value, and Ld2 is the second characteristic pixel point brightness value;

[0122] Step S233: Use an edge detection algorithm to perform edge monitoring on the sample image quality analysis image to obtain the edge intensity value corresponding to the sample image quality analysis image;

[0123] Step S234: Calculate the edge intensity value, the original image contrast and the original resolution value to obtain the image quality coefficient corresponding to the sample image quality analysis image;

[0124] The image quality coefficient corresponding to the sample image quality analysis image is calculated. The specific formula is as follows:

[0125] ;

[0126] Among them, Hzx is the image quality coefficient, Byq is the edge strength value, Dbd is the original image contrast, and Fbl is the original resolution value;

[0127] Step S24: performing image quality analysis on the first terminal image to the a-th terminal image respectively to obtain image quality coefficients corresponding to the first terminal image to the a-th terminal image respectively;

[0128] Step S25: defining the image quality coefficients corresponding to the first terminal image to the a-th terminal image as image quality data;

[0129] Step S3: processing each terminal image into an encrypted terminal image according to the image privacy data and the image quality data, obtaining a plurality of encrypted terminal images, transmitting the encrypted terminal images, and optimizing the image quality of each terminal image according to the image quality data;

[0130] Step S31: Acquire image privacy data, and acquire the first image privacy area to the ath image privacy area respectively according to the image privacy data;

[0131] Step S32: privacy encrypting the first image privacy area to the a-th image privacy area to obtain the first image privacy encryption area to the a-th image privacy encryption area, respectively covering the first image privacy area to the a-th image privacy encryption area with the first image privacy encryption area to the a-th image privacy encryption area, obtaining the first encrypted terminal image as the a-th encrypted terminal image, and transmitting the first encrypted terminal image as the a-th encrypted terminal image to the receiving end user;

[0132] The step S32 further includes the following specific steps:

[0133] Step S321: Generate a byte sequence of 128 bits in length as a generated key using a random number generator and transmit the generated key to a data decryption unit;

[0134] Step S322: Convert the first image privacy area into the first image privacy area Unicode encoding through a Unicode decoder, and then convert the first image privacy area Unicode encoding into the first image privacy area binary encoding. Divide the first image privacy area binary encoding into data packets of equal length, with each data packet having a length of 128 bits, corresponding to a 128-bit byte sequence key;

[0135] Step S323: Use the previous data packet of the current data packet as the initial vector, perform an exclusive OR operation (resulting in 1 for different bits and 0 for the same bits) on the current data packet and the initial vector, and then encrypt the result of the exclusive OR operation using the generated key. Use the encrypted data as the initial vector for the next data packet. Repeat the above process, and use the last encrypted data as the initial vector for the first data packet to complete the encryption of all data packets, thereby achieving the encryption of the first image privacy area and obtaining the first image privacy encrypted area;

[0136] Step S324: Encrypt the second image privacy area to the a-th image privacy area respectively to obtain the second image privacy encrypted area to the a-th image privacy encrypted area;

[0137] Step S33: Obtain the image quality data, and obtain the image quality coefficients corresponding to the first terminal image to the a-th terminal image according to the image quality data;

[0138] Step S34: Adjust the image quality of the first terminal image;

[0139] In the said step S34, the following specific steps are further included:

[0140] Step S341: Obtain the calibrated screen contrast ratio, calibrated screen edge intensity value, and calibrated screen resolution value of the image display device corresponding to the first terminal image respectively;

[0141] Step S342: Calculate the screen display quality coefficient corresponding to the image display device through the calibrated screen contrast ratio, calibrated screen edge intensity value, and calibrated screen resolution value;

[0142] Calculate the screen display quality coefficient, and the specific formula is as follows:

[0143] ;

[0144] Among them, Pzx is the screen display quality coefficient, Pyq is the calibrated screen edge intensity value, Pbd is the calibrated screen contrast ratio, and Pbl is the calibrated screen resolution value;

[0145] Step S343: Obtain the image quality coefficient corresponding to the first terminal image;

[0146] Step S344: When the image quality coefficient is greater than or equal to the screen display quality coefficient, directly output the first terminal image through the display device;

[0147] Step S345: When the image quality coefficient is less than the screen display quality coefficient, directly optimize the display quality of the first terminal image until the image quality coefficient is equal to the screen display quality coefficient;

[0148] Step S345: Adjust the display quality of the second terminal image to the a-th terminal image respectively;

[0149] In this application, if there are corresponding calculation formulas, the above calculation formulas are all dimensionless and take their numerical values for calculation. For coefficients such as weight coefficients and proportionality coefficients in the formulas, the sizes set are for obtaining a result value by quantifying each parameter. Regarding the sizes of the weight coefficients and proportionality coefficients, as long as they do not affect the proportional relationship between the parameters and the result value.

[0150] Embodiment 2

[0151] Please refer to Figure 2 , based on another concept of the same invention, a system for acquiring image data of an image processing chip is proposed. The system for acquiring image data includes a privacy data module, an image quality module, a data processing module, and a server. The privacy data module, the image quality module, and the data processing module are respectively connected to the server, and the server controls the privacy data module, the image quality module, and the data processing module respectively;

[0152] The privacy data module respectively acquires and collects multiple terminal images, conducts privacy analysis on each terminal image respectively, and obtains the image privacy area corresponding to each terminal image to obtain image privacy data;

[0153] Acquire the images stored in the image processing terminal where the image processing chip is located, and name the stored images as the first terminal image to the a-th terminal image according to the saving time respectively;

[0154] It should be noted here that:

[0155] Here, a is the numerical value corresponding to the number of terminal images, and a is an integer greater than 0. The image processing terminal involved here is specifically an electronic device for receiving, processing, and displaying images. The image processing terminal involved here includes but is not limited to a PC, a tablet, and a mobile phone;

[0156] Mark the privacy areas of the first terminal image to the a-th terminal image to obtain the first image privacy data to the a-th image privacy area;

[0157] Specifically as follows:

[0158] Randomly select one terminal image from the first terminal image to the a-th terminal image as the sample privacy image;

[0159] Please refer to Figure 3 , perform layer scanning on the sample privacy image through OCR technology, and divide the image content in the sample privacy image into the first image content and the second image content;

[0160] It should be noted here that:

[0161] In this application, the first image content is the corresponding text content in the sample privacy image, and the second image content is the corresponding image content in the sample privacy image;

[0162] Mark the privacy content of the first image content as follows:

[0163] Set multiple privacy keywords respectively, use text recognition algorithms to perform text analysis on the first image content, mark the areas in the first image content with the same text as any privacy keyword as the image privacy areas, and mark the image privacy areas;

[0164] It should be noted here that:

[0165] The privacy keywords involved here include but are not limited to names, ID numbers, and mobile phone numbers;

[0166] Mark the privacy content of the second image content as follows:

[0167] Obtain multiple images with privacy content as sample privacy image data through web crawler technology, and label the privacy areas corresponding to each image in the sample privacy image data through manual annotation to obtain sample privacy image annotation data;

[0168] Divide the multiple images in the sample privacy image data into an image training set and an image test set according to the image division ratio;

[0169] Create an image recognition model through an existing artificial intelligence platform, use the image training set to perform image recognition training on the image recognition model until the image recognition model has completed image recognition training for each image in the image training set;

[0170] Use the image test set to perform training tests on the image recognition model, obtain the recognition accuracy, obtain the current recognition accuracy, obtain the benchmark recognition accuracy. If the current recognition accuracy is greater than or equal to the benchmark recognition accuracy, the training of the image recognition model is completed to obtain the privacy content recognition model. If the current recognition accuracy is less than the benchmark recognition accuracy, use the image training set to continue training the image recognition model until the current recognition accuracy is greater than or equal to the benchmark recognition accuracy;

[0171] It should be noted here that:

[0172] In this application, the image division ratio involved here is 3:7, and the reference recognition accuracy involved here is 95%;

[0173] Use the privacy content recognition model to perform privacy recognition on the second image content, and mark the recognized privacy area as the image privacy area;

[0174] Repeat the process of obtaining the image privacy area corresponding to the sample privacy image, and perform image privacy area acquisition on the first terminal image to the a-th terminal image respectively to obtain the first image privacy area to the a-th image privacy area;

[0175] Define the first image privacy area to the a-th image privacy area as image privacy data;

[0176] The privacy data module acquires the image privacy data and transports it to the image quality module and the data processing module;

[0177] The image quality module performs image quality analysis on each terminal image according to the image privacy data, and obtains multiple image quality coefficients to obtain image quality data;

[0178] Acquire the image privacy data, and respectively acquire the first terminal image to the a-th terminal image according to the image privacy data;

[0179] Randomly select an image from the first terminal image to the a-th terminal image as the sample image quality analysis image;

[0180] Perform image quality analysis on the sample image quality analysis image to obtain the image quality coefficient corresponding to the sample image quality analysis image;

[0181] Obtain the original resolution value corresponding to the sample image quality analysis image;

[0182] Specifically as follows:

[0183] Randomly select multiple pixel distribution rows in the sample image quality analysis image, respectively obtain the pixel point values corresponding to each pixel distribution row to obtain multiple horizontal pixel point quantity values, and perform an average calculation on the obtained multiple horizontal pixel point quantity values to obtain the horizontal pixel point average quantity value;

[0184] Randomly select multiple pixel distribution columns in the sample image quality analysis image, respectively obtain the pixel point values corresponding to each pixel distribution column to obtain multiple vertical pixel point quantity values, and perform an average calculation on the obtained multiple vertical pixel point quantity values to obtain the vertical pixel point average quantity value;

[0185] Calculate the product of the average number of horizontal pixel points and the average number of vertical pixel points to obtain the original resolution value corresponding to the sample image quality analysis image;

[0186] Obtain the original image contrast corresponding to the sample image quality analysis image as follows:

[0187] Randomly select m monitoring pixel points in the sample image quality analysis image, and obtain the brightness values corresponding to each pixel point respectively to get multiple pixel point brightness values;

[0188] Compare the magnitudes of the obtained multiple pixel point brightness values, mark the pixel point brightness value with the largest value as the first characteristic pixel point brightness value, and mark the pixel point brightness value with the smallest value as the second characteristic pixel point brightness value;

[0189] Calculate the original image contrast corresponding to the sample image quality analysis image from the first characteristic pixel point brightness value and the second characteristic pixel point brightness value;

[0190] Calculate the original image contrast corresponding to the sample image quality analysis image, and the specific formula configuration is as follows:

[0191] ;

[0192] Among them, Dbd is the original image contrast corresponding to the sample image quality analysis image, Ld1 is the first characteristic pixel point brightness value, and Ld2 is the second characteristic pixel point brightness value;

[0193] Use the edge detection algorithm to perform edge monitoring on the sample image quality analysis image to obtain the edge intensity value corresponding to the sample image quality analysis image;

[0194] Calculate the image quality coefficient corresponding to the sample image quality analysis image from the edge intensity value, the original image contrast, and the original resolution value;

[0195] Calculate the image quality coefficient corresponding to the sample image quality analysis image, and the specific formula is as follows:

[0196] ;

[0197] Among them, Hzx is the image quality coefficient, Byq is the edge intensity value, Dbd is the original image contrast, and Fbl is the original resolution value;

[0198] Repeat the process of performing image quality analysis on the sample image quality analysis image, and perform image quality analysis on the first terminal image to the a-th terminal image respectively to obtain the image quality coefficients corresponding to the first terminal image to the a-th terminal image;

[0199] Define the image quality coefficients corresponding to the first terminal image to the a-th terminal image as image quality data;

[0200] The image quality module obtains the image quality data and conveys it to the data processing module;

[0201] The data processing module processes each terminal image into an encrypted terminal image according to the image privacy data and the image quality data, obtains multiple encrypted terminal images, transmits the encrypted terminal images, and optimizes the image quality of each terminal image according to the image quality data;

[0202] Obtain the image privacy data, and respectively obtain the first image privacy area to the a-th image privacy area according to the image privacy data;

[0203] Perform privacy encryption on the first image privacy area to the a-th image privacy area to obtain the first image privacy encrypted area to the a-th image privacy encrypted area, and cover the first image privacy area to the a-th image privacy area through the first image privacy encrypted area to the a-th image privacy encrypted area respectively, to obtain the first encrypted terminal image to the a-th encrypted terminal image, and transmit the first encrypted terminal image to the a-th encrypted terminal image to the receiving end user;

[0204] Specifically as follows:

[0205] Use a random number generator to generate a byte sequence of 128 bits as the generation key and convey the generation key to the data decryption unit;

[0206] Convert the first image privacy area into the Unicode encoding of the first image privacy area through a Unicode decoder, and then convert the Unicode encoding of the first image privacy area into the binary encoding of the first image privacy area. Divide the binary encoding of the first image privacy area into data groups of equal length, and the length of each data group is 128 bits, corresponding to the 128-bit byte sequence key;

[0207] It should be noted here that: if the length of the binary encoding of the first image privacy area is not an integer multiple of the data group length, then fill the high bits of the binary encoding of the patient pre-consultation data with 0 until the binary encoding of the first image privacy area is an integer multiple of the data group length;

[0208] Use the previous data grouping of the current data grouping as the initial vector, perform an exclusive OR operation (resulting in 1 for different values and 0 for the same values) between the current data grouping and the initial vector, then encrypt the result of the exclusive OR operation using the generated key, use the encrypted data as the initial vector for the next data grouping, repeat the above process, and use the last encrypted data as the initial vector for the first data grouping to complete the encryption of all data groupings, thereby achieving the encryption of the privacy area of the first image and obtaining the first image privacy encryption area;

[0209] Encrypt the privacy areas of the second image to the a-th image respectively to obtain the second image privacy encryption area to the a-th image privacy encryption area;

[0210] It should be noted here that:

[0211] In this application, the collected terminal image is processed into an encrypted terminal image for transmission, and the user who receives the encrypted terminal image can decrypt the encrypted terminal image through the key. The encryption algorithm involved here is the AES symmetric encryption algorithm, and the encryption key and the decryption key are the same key;

[0212] Obtain the image quality data, and obtain the image quality coefficients corresponding to the first terminal image to the a-th terminal image according to the image quality data;

[0213] Perform image quality adjustment on the first terminal image;

[0214] Specifically as follows:

[0215] Respectively obtain the calibrated screen contrast ratio, calibrated screen edge intensity value, and calibrated screen resolution value of the image display device corresponding to the first terminal image;

[0216] It should be noted here that:

[0217] The calibrated screen contrast ratio, calibrated screen edge intensity value, and calibrated screen resolution value involved here are respectively the maximum values of the screen contrast ratio, screen edge intensity value, and screen resolution corresponding to the image display device;

[0218] Calculate the calibrated screen contrast ratio, calibrated screen edge intensity value, and calibrated screen resolution value to obtain the screen display quality coefficient corresponding to the image display device;

[0219] Calculate the screen display quality coefficient, and the specific formula is as follows:

[0220] ;

[0221] Among them, Pzx is the screen display quality coefficient, Pyq is the calibrated screen edge intensity value, Pbd is the calibrated screen contrast ratio, and Pbl is the calibrated screen resolution value;

[0222] Obtain the image quality coefficient corresponding to the first terminal image;

[0223] When the image quality coefficient is greater than or equal to the screen display quality coefficient, directly output the first terminal image through the display device;

[0224] When the image quality coefficient is less than the screen display quality coefficient, directly optimize the display quality of the first terminal image until the image quality coefficient is equal to the screen display quality coefficient;

[0225] Adjust the display quality of the second terminal image to the a-th terminal image respectively.

[0226] The preferred embodiments of the present invention disclosed above are only used to help illustrate the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to the specific embodiments. Obviously, many modifications and changes can be made according to the content of this specification. These embodiments are selected and specifically described in this specification to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and utilize the present invention well.

Claims

1. An image data acquisition method for an image processing chip, characterized in that: include: Step S1: Acquire multiple collected terminal images, perform privacy analysis on each terminal image respectively, and acquire the image privacy area corresponding to each terminal image to obtain image privacy data; Step S2: Perform image quality analysis on each terminal image according to the image privacy data, and obtain multiple image quality coefficients to obtain image quality data; Step S21: Obtain image privacy data, and obtain the first terminal image to the a-th terminal image respectively according to the image privacy data; Step S22: Randomly select an image from the first terminal image to the a-th terminal image as a sample image quality analysis image; Step S23: Perform image quality analysis on the sample image quality analysis image to obtain the image quality coefficient corresponding to the sample image quality analysis image; Step S231: Obtain the original resolution value corresponding to the sample image quality analysis image; Step S232: The original image contrast corresponding to the image is obtained; step S233: edge detection algorithm is used to monitor the edge of the sample image quality analysis image to obtain the edge strength value corresponding to the sample image quality analysis image; step S234: the edge strength value, the original image contrast and the original resolution value are summed to obtain the image quality coefficient corresponding to the sample image quality analysis image; step S24: image quality analysis is performed on the first terminal image to the a-th terminal image respectively to obtain the image quality coefficients corresponding to the first terminal image to the a-th terminal image respectively; step S25: the image quality coefficients corresponding to the first terminal image to the a-th terminal image are defined as image quality data; Step S3: processing each terminal image into an encrypted terminal image according to the image privacy data and the image quality data, obtaining a plurality of encrypted terminal images, transmitting the encrypted terminal images, and optimizing the image quality of each terminal image according to the image quality data; Step S33: acquiring image quality data, and acquiring image quality coefficients corresponding to the first terminal image to the ath terminal image respectively according to the image quality data; Step S34: adjusting the image quality of the first terminal image; Step S341: respectively obtaining the calibrated screen contrast, calibrated screen edge strength value and calibrated screen resolution value of the image display device corresponding to the first terminal image; Step S342: summing up the calibrated screen contrast, calibrated screen edge strength value and calibrated screen resolution value to obtain the screen display quality coefficient corresponding to the image display device; Step S343: obtaining the image quality coefficient corresponding to the first terminal image; Step S344: when the image quality coefficient is greater than or equal to the screen display quality coefficient, directly outputting the first terminal image through the display device; Step S345: when the image quality coefficient is less than the screen display quality coefficient, directly optimizing the display quality of the first terminal image until the image quality coefficient is equal to the screen display quality coefficient; Step S346: respectively adjusting the display quality of the second terminal image to the ath terminal image.

2. The image data acquisition method of an image processing chip according to claim 1, characterized in that: The step S1 further includes the following specific steps: Step S11: acquiring images stored in the image processing terminal where the image processing chip is located, and naming the stored images as first terminal image to ath terminal image according to the storage time; Step S12: marking the privacy area of ​​the first terminal image to the a-th terminal image to obtain the privacy area of ​​the first image privacy data to the a-th image; Step S13: randomly selecting a terminal image from the first terminal image to the ath terminal image as a sample privacy image; Step S14: performing privacy marking on the sample privacy image to obtain the image privacy area; Step S15: acquiring image privacy areas from the first terminal image to the a-th terminal image respectively, to obtain the first image privacy area to the a-th image privacy area; Step S16: defining the first image privacy area to the ath image privacy area as image privacy data.

3. The image data acquisition method of an image processing chip according to claim 2, characterized in that: The step S14 further includes the following specific steps: Step S141: performing layer scanning on the sample privacy image by using OCR technology, and dividing the image content in the sample privacy image into first image content and second image content; Step S142: setting a plurality of privacy keywords respectively, performing text analysis on the first image content using a text recognition algorithm, marking an area in the first image content that has the same text as any privacy keyword as an image privacy area, and marking the image privacy area; Step S143: marking the second image content with privacy content to obtain an image privacy area corresponding to the second image content.

4. The image data acquisition method of an image processing chip according to claim 3, characterized in that: The step S143 further includes the following specific steps: Step S1431: using a web crawler technology to obtain multiple images with private content as sample private image data, and manually annotating the private area corresponding to each image of the sample private image data to obtain sample private image annotated data; Step S1432: dividing the multiple images in the sample private image data into an image training set and an image test set according to the image division ratio; Step S1433: creating an image recognition model through an existing artificial intelligence platform, and performing image recognition training on the image recognition model using the image training set, until the image recognition model completes image recognition training for each image in the image training set; Step S1434: Use the image test set to train and test the image recognition model, and obtain the recognition accuracy to obtain the current recognition accuracy and the benchmark recognition accuracy. If the current recognition accuracy is greater than or equal to the benchmark recognition accuracy, the image recognition model training is completed and the private content recognition model is obtained. If the current recognition accuracy is less than the benchmark recognition accuracy, the image training set is used to continue training the image recognition model until the current recognition accuracy is greater than or equal to the benchmark recognition accuracy. Step S1435: Use the privacy content recognition model to perform privacy recognition on the second image content, and mark the recognized privacy area as the image privacy area.

5. The image data acquisition method of an image processing chip according to claim 1, characterized in that: The step S231 further includes the following specific steps: Step S2311: randomly selecting a plurality of pixel distribution lines in the sample image quality analysis image, respectively obtaining the pixel point value corresponding to each pixel distribution line, obtaining a plurality of horizontal pixel point quantity values, and averaging the obtained plurality of horizontal pixel point quantity values ​​to obtain a horizontal pixel point average quantity value; Step S2312: randomly selecting a plurality of pixel distribution columns in the sample image quality analysis image, respectively obtaining the pixel point value corresponding to each pixel distribution column, obtaining a plurality of longitudinal pixel point quantity values, and averaging the obtained plurality of longitudinal pixel point quantity values ​​to obtain a longitudinal pixel point average quantity value; Step S2313: Calculate the product of the average number of horizontal pixels and the average number of vertical pixels to obtain the original resolution value corresponding to the sample image quality analysis image.

6. The image data acquisition method of an image processing chip according to claim 1, characterized in that: The step S232 further includes the following specific steps: Step S2321: randomly selecting m monitoring pixels from the sample image quality analysis image, and acquiring the brightness value corresponding to each pixel to obtain a plurality of pixel brightness values; Step S2322: Compare the obtained multiple pixel brightness values, mark the pixel brightness value with the largest value as the first characteristic pixel brightness value, and mark the pixel brightness value with the smallest value as the second characteristic pixel brightness value; Step S2323: calculating the brightness value of the first characteristic pixel and the brightness value of the second characteristic pixel to obtain the original image contrast corresponding to the sample image quality analysis image; Calculate the contrast of the original image corresponding to the sample image quality analysis image. The specific formula configuration is as follows: ； Wherein, Dbd is the original image contrast corresponding to the sample image quality analysis image, Ld1 is the brightness value of the first characteristic pixel, and Ld2 is the brightness value of the second characteristic pixel.

7. The image data acquisition method of an image processing chip according to claim 1, characterized in that: The step S3 further includes the following specific steps: Step S31: Acquire image privacy data, and acquire the first image privacy area to the ath image privacy area respectively according to the image privacy data; Step S32: privacy encrypt the first image privacy area to the a-th image privacy area to obtain the first image privacy encryption area to the a-th image privacy encryption area, and respectively cover the first image privacy area to the a-th image privacy encryption area with the first image privacy encryption area to the a-th image privacy encryption area to obtain the first encrypted terminal image as the a-th encrypted terminal image, and transmit the first encrypted terminal image as the a-th encrypted terminal image to the receiving end user.

8. The image data acquisition method of an image processing chip according to claim 7, characterized in that: The step S32 further includes the following specific steps: Step S321: Generate a byte sequence of 128 bits in length as a generated key using a random number generator and transmit the generated key to a data decryption unit; Step S322: converting the first image privacy area into the first image privacy area Unicode code by a Unicode decoder, and then converting the first image privacy area Unicode code into the first image privacy area binary code, and dividing the first image privacy area binary code into data groups of equal length, each data group has a length of 128 bits, corresponding to a 128-bit byte sequence key; Step S323: the previous data group of the current data group is used as the initial vector, the current data group and the initial vector are subjected to an XOR operation (different data are 1, and identical data are 0), and the result of the XOR operation is encrypted using the generated key, and the encrypted data is used as the initial vector of the next data group. The above process is repeated, and the last encrypted data is used as the initial vector of the first data group. The encryption of all data groups is completed, and the encryption of the first image privacy area is realized, thereby obtaining the first image privacy encryption area; Step S324: Encrypt the second image privacy area to the a-th image privacy area respectively to obtain the second image privacy encryption area to the a-th image privacy encryption area.

9. An image data acquisition system for an image processing chip, applicable to an image data acquisition method for an image processing chip according to any one of claims 1 to 8, characterized in that: The specific working process of the image data acquisition system is as follows: Privacy data module: used to acquire and collect multiple terminal images, perform privacy analysis on each terminal image, and acquire the image privacy area corresponding to each terminal image to obtain image privacy data; Image quality module: used to analyze the image quality of each terminal image according to the image privacy data, and obtain multiple image quality coefficients to obtain image quality data; Data processing module: used to process each terminal image into an encrypted terminal image according to the image privacy data and the image quality data, obtain multiple encrypted terminal images, transmit the encrypted terminal images, and optimize the image quality of each terminal image according to the image quality data.

Images