CN112767327A - Image quality management system and method based on neural network - Google Patents
Image quality management system and method based on neural network Download PDFInfo
- Publication number
- CN112767327A CN112767327A CN202110024819.1A CN202110024819A CN112767327A CN 112767327 A CN112767327 A CN 112767327A CN 202110024819 A CN202110024819 A CN 202110024819A CN 112767327 A CN112767327 A CN 112767327A
- Authority
- CN
- China
- Prior art keywords
- image
- signal
- color
- evaluation
- detection
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/90—Determination of colour characteristics
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30168—Image quality inspection
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Quality & Reliability (AREA)
- Mobile Radio Communication Systems (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses an image quality management system and method based on a neural network, which relates to the technical field of image quality management and solves the technical problem that the quality of an image is reduced due to the fact that the image cannot be evaluated in the prior art, an image evaluation unit is used for analyzing original data of the image so as to evaluate the image, the number of pixels of the image to be evaluated in the horizontal direction and the number of pixels of the image to be evaluated in the vertical direction are obtained in real time, any pixel is obtained in the image to be evaluated, the pixel is substituted into a calculation formula to obtain the brightness mean value of the image to be evaluated, the pixel gray value standard difference of the image to be evaluated and the average gradient of the image to be evaluated, the brightness mean value of the image, the pixel gray value standard difference of the image and the average gradient of the image are calculated to evaluate the image, and the quality of the image is, the method brings convenience to image quality management, thereby improving the working efficiency.
Description
Technical Field
The invention relates to the technical field of image quality management, in particular to an image quality management system and method based on a neural network.
Background
With the continuous improvement of scientific technology. Computer science is becoming mature, strong functions of the computer science are deeply known, the computer science has entered various fields of human society and plays more and more important roles, the computer science is the inevitable requirement of personal application computing management daily business, and the image management system is used for carrying out detailed management on added images and image management, so that people can quickly find the images which are wanted by themselves. Today, computers are already very inexpensive, but have made significant advances in performance.
At that time, in the prior art, the image could not be evaluated, resulting in a reduction in the quality of the image, thereby reducing the work efficiency.
Disclosure of Invention
The invention aims to provide an image quality management system and method based on a neural network, wherein image original data are analyzed through an image evaluation unit, so that an image is evaluated, the number of pixels of the image to be evaluated in the horizontal direction and the number of pixels of the image to be evaluated in the vertical direction are obtained in real time, the number of pixels of the image in the horizontal direction and the number of pixels of the image in the vertical direction are respectively marked as X and Y, and then the resolution of the image to be evaluated is marked as X multiplied by Y; any pixel point is obtained in the image to be evaluated, the pixel point is substituted into a calculation formula to obtain the brightness mean value of the image to be evaluated, the pixel gray value standard deviation of the image to be evaluated and the average gradient of the image to be evaluated, the brightness mean value of the image, the pixel gray value standard deviation of the image and the average gradient of the image are subjected to evaluation coefficient calculation, the image is evaluated, the quality of the image is improved, convenience is brought to image quality management, and therefore working efficiency is improved.
The purpose of the invention can be realized by the following technical scheme:
the image quality management system based on the neural network comprises an image detection unit, an image evaluation unit, a color regulation unit, a test evaluation unit, a selection storage unit, an image management platform, a registration login unit and a database;
the image evaluation unit is used for analyzing image raw data so as to evaluate an image, wherein the image raw data comprises a brightness mean value of the image, a pixel gray value standard deviation of the image and an average gradient of the image, the image is marked as i, i is 1, 2, … …, n, n is a positive integer, and the specific analysis and evaluation process is as follows:
the method comprises the steps of firstly, acquiring the number of pixels of an image to be evaluated in the horizontal direction and the number of pixels in the vertical direction in real time, marking the number of pixels of the image in the horizontal direction and the number of pixels in the vertical direction as X and Y respectively, and then marking the resolution of the image to be evaluated as X multiplied by Y;
step two, obtaining any pixel point in the image to be evaluated, marking the pixel point as F (s, e), substituting the pixel point into a calculation formula to obtain the brightness mean value of the image to be evaluated, marking the brightness mean value of the image to be evaluated as LJi, and the calculation formula is as follows:
wherein, alpha is an error correction factor and takes the value of 2.3654512;
step three, substituting the pixel points and the brightness mean value of the image to be evaluated into a calculation formula to obtain the pixel gray value standard deviation of the image to be evaluated, marking the pixel gray value standard deviation of the image to be evaluated as BZi, wherein the calculation formula is as follows:
and substituting the pixel points into a calculation formula to obtain the average gradient of the image to be evaluated, marking the average gradient of the image to be evaluated as TDi, wherein the calculation formula is as follows:
wherein, β is a fixed threshold value set, and the value is 5.63212302;
fourthly, calculating the evaluation coefficient of the brightness mean value of the image, the standard deviation of the pixel gray value of the image and the average gradient of the image, wherein the calculation formula is
Wherein PGi is the evaluation coefficient of the image, a1, a2 and a3 are all proportionality coefficients, and a1 > a2 > a3 > 0;
step five, comparing the evaluation coefficient of the image with an evaluation coefficient threshold value:
if the evaluation coefficient of the image is larger than or equal to the evaluation coefficient threshold value, judging that the corresponding image is qualified for evaluation, generating an evaluation qualified signal and sending the evaluation qualified signal to the image management platform;
and if the evaluation coefficient of the image is less than the evaluation coefficient threshold value, judging that the corresponding image is unqualified in evaluation, generating an unqualified evaluation signal and sending the unqualified evaluation signal to a mobile phone terminal of an operator, and after receiving the unqualified evaluation signal, the operator performs numerical verification on the brightness mean value of the image, the pixel gray value standard deviation of the image and the average gradient of the image.
Further, the image management platform generates an image detection signal after receiving the image evaluation qualified signal, and sends the image detection signal to the image detection unit, the image detection unit analyzes the existing image data after receiving the image detection signal, so as to detect the image, the existing image data includes compressed data, line-pair data and cycle data, the compressed data is the number of characters reduced in the encoding and compressing process of the image data, the line-pair data is the number of black and white lines in the image with equal width per millimeter, the cycle data is the number of image space cycles per millimeter, and the specific analysis and detection process is as follows:
step S1: acquiring the number of characters reduced in the encoding and compressing process of the image data, and marking the number of characters reduced in the encoding and compressing process of the image data as ZFi;
step S2: acquiring black and white line logarithm in the image with the same width per millimeter, and marking the black and white line logarithm in the image with the same width per millimeter as DSi;
step S3: acquiring the number of image space cycles per millimeter, and marking the number of image space cycles per millimeter as ZQi;
step S4: by calculation formula
Acquiring detection coefficients JCi of the image, wherein v1, v2 and v3 are all proportionality coefficients, and v1 is more than v2 is more than v3 is more than 0;
step S5: compare the detection coefficient JCi of the image to a detection coefficient threshold of the image:
if the detection coefficient JCi of the image is not less than the detection coefficient threshold of the image, judging that the corresponding image is qualified for detection, generating an image qualified detection signal and sending the image qualified detection signal to an image management platform, and after receiving the image qualified detection signal, the image management platform generates an image test signal and sends the image test signal to a test evaluation unit;
and if the detection coefficient JCi of the image is less than the detection coefficient threshold of the image, judging that the corresponding image is unqualified, generating an unqualified image detection signal and sending the unqualified image detection signal to the mobile phone terminal of the operator.
Further, the test evaluation unit is used for performing test evaluation on the image so as to detect the image, and the specific test detection process is as follows:
step SS 1: randomly selecting ten testers at different age ends, observing the images by the ten testers, counting the content information quantity in the images by the ten testers after the observation is finished, calculating the average content information quantity of the images by averaging, and marking the average content information quantity as PJSi;
step SS 2: then acquiring an image gray distribution set P ═ (P1, P2, … …, Pn) according to the gray distribution of the image, then acquiring the probability of the gray value appearing in the image, and marking the probability of the gray value appearing in the image as P (t);
step SS 3: acquiring an entropy value of a corresponding image through a calculation formula, wherein the calculation formula is as follows:
wherein Si is an entropy value of a corresponding image, and 64 is a fixed gray level threshold value of the image;
step SS 4: comparing the entropy value of the corresponding image with a threshold value of entropy value:
if the entropy value of the corresponding image is larger than or equal to the threshold value of the entropy value, judging that the corresponding image is qualified in the test, generating a qualified test signal and sending the qualified test signal to the image management platform, and after receiving the qualified test signal, the image management platform generates a color detection signal and sends the color detection signal to the color adjusting unit;
and if the entropy value of the corresponding image is less than the threshold value of the entropy value, judging that the corresponding image test is unqualified, generating a test unqualified signal and sending the test unqualified signal to the mobile phone terminal of the operator.
Further, after receiving the color detection signal, the color adjustment unit analyzes the color data of the image to adjust the color of the image, where the color data of the image is the total number of colors in the color of the image, the ratio of the number of colors in the warm color system to the number of colors in the cold color system, and the saturation of the colors, and the specific analysis adjustment process is as follows:
step L1: acquiring the total number of colors in the colors of the image, and marking the total number of the colors in the colors of the image as ZSLI;
step L2: acquiring the color quantity ratio of a warm color system and a cold color system in the colors of the image, and marking the color quantity ratio of the warm color system and the cold color system in the colors of the image as SBZi;
step L3: acquiring the color saturation of the image, and marking the color saturation of the image as BGDi;
step (ii) ofL4: by calculation formula
Acquiring color adjustment coefficients TJi of the image, wherein s1, s2 and s3 are all proportionality coefficients, and s1 is more than s2 is more than s3 is more than 0;
step L5: compare the color-toning coefficient TJi for an image to a color-toning coefficient threshold for the image:
if the color adjustment coefficient TJi of the image is not less than the color adjustment coefficient threshold of the image, judging that the corresponding image does not need to be adjusted, generating a color detection qualified signal and sending the color detection qualified signal to the image management platform;
and if the color adjustment coefficient TJi of the image is less than the color adjustment coefficient threshold value of the image, judging that the corresponding image needs to be adjusted, generating a unqualified color detection signal and sending the unqualified color detection signal to the mobile phone terminal of the administrator.
Further, after receiving the evaluation qualified signal, the image detection qualified signal, the test qualified signal and the color detection qualified signal, the image management platform generates an image storage signal and sends the image storage signal to the selective storage unit, and after receiving the image storage signal, the selective storage unit stores the image, and the specific storage process is as follows:
step T1: setting an independent storage area in a database, and marking the independent storage area as o, o is 1, 2, … …, m is a positive integer;
step T2: acquiring a space memory of an independent storage area, marking the space memory of the independent storage area as NCo, sequencing the independent storage area according to the numerical value of the space memory of the independent storage area, and sending the sequenced independent storage area to an image management platform;
step T3: sorting the images to be stored according to the size of the memory data of the images to be stored, and sending the sorted images to be stored to an image management platform;
step T4: after receiving the sequenced independent storage areas and the sequenced images to be stored, the image management platform matches the independent storage areas and the images to be stored one by one according to the sequence, then generates a protection demand signal and sends the protection demand signal to a mobile phone terminal of a manager, after the manager receives the protection demand signal, if the corresponding image needs to be protected, a determined protection signal is generated and sent to the image management platform, and if the corresponding image does not need to be protected, a signal which does not need to be protected is generated and sent to the image management platform;
step T5: and after receiving the determined protection signal, the image management platform sets a single storage path for the corresponding protection image, and limits the use times of the single storage path, wherein the single storage path represents the only storage path for setting the corresponding protection image.
Further, the registration login unit is used for a manager and an operator to submit manager information and operator information through a mobile phone terminal, and the manager information and the operator information which are successfully registered are sent to the database for storage, the manager information comprises the name, the age, the time of entry and the mobile phone number of the real name authentication of the person, and the operator information comprises the name, the age, the time of entry and the mobile phone number of the real name authentication of the person.
Further, the image quality management method based on the neural network comprises the following specific steps:
step P1: image evaluation, namely analyzing original image data through an image evaluation unit so as to evaluate the image, and then sending an evaluation qualified signal to an image management platform;
step P2: the image detection is to analyze the existing data of the image after receiving the image detection signal through the image detection unit so as to detect the image, and then to send a detection qualified signal to the image management platform;
step P3: image test evaluation, namely performing test evaluation on the image through a test evaluation unit so as to detect the image, and then sending a test qualified signal to an image management platform;
step P4: the image color adjustment is to analyze the color data of the image after receiving the color detection signal through the color adjustment unit so as to adjust the color of the image, and then the color detection qualified signal is sent to the image management platform;
step P5: and selecting and storing the image, wherein the image management platform generates an image storage signal and sends the image storage signal to the selection storage unit after receiving the evaluation qualified signal, the image detection qualified signal, the test qualified signal and the color detection qualified signal, and then stores the image after receiving the image storage signal through the selection storage unit.
Compared with the prior art, the invention has the beneficial effects that:
1. in the invention, the original data of the image is analyzed by an image evaluation unit, so that the image is evaluated, the number of pixels of the image to be evaluated in the horizontal direction and the number of pixels of the image to be evaluated in the vertical direction are obtained in real time, the number of pixels of the image in the horizontal direction and the number of pixels of the image in the vertical direction are respectively marked as X and Y, and then the resolution of the image to be evaluated is marked as X multiplied by Y; obtaining any pixel point in an image to be evaluated, marking the pixel point as F (s, e), substituting the pixel point into a calculation formula to obtain a brightness mean value of the image to be evaluated, a pixel gray value standard difference of the image to be evaluated and an average gradient of the image to be evaluated, carrying out evaluation coefficient calculation on the brightness mean value of the image, the pixel gray value standard difference of the image and the average gradient of the image, judging that the corresponding image is qualified for evaluation if an evaluation coefficient of the image is more than or equal to an evaluation coefficient threshold value, generating an evaluation qualified signal and sending the evaluation qualified signal to an image management platform; if the evaluation coefficient of the image is smaller than the evaluation coefficient threshold value, judging that the corresponding image is unqualified in evaluation, generating an unqualified evaluation signal and sending the unqualified evaluation signal to a mobile phone terminal of an operator, and after receiving the unqualified evaluation signal, the operator performs numerical verification on the brightness mean value of the image, the pixel gray value standard deviation of the image and the average gradient of the image; the image is evaluated, so that the quality of the image is improved, convenience is brought to image quality management, and the working efficiency is improved;
2. in the invention, the image is tested by performing test evaluation on the image through a test evaluation unit, ten testers with different ages are randomly selected and observed by the ten testers, after the observation is finished, the ten testers count the content information quantity in the image, the average quantity of the content information of the image is obtained by averaging, then an image gray distribution set P (P1, P2, … … and Pn) is obtained according to the gray distribution of the image, then the probability of the gray value appearing in the image is obtained, and the probability of the gray value appearing in the image is marked as P (t); acquiring an entropy value of a corresponding image through a calculation formula, judging that the corresponding image is qualified in a test if the entropy value of the corresponding image is larger than or equal to a threshold value of the entropy value, generating a qualified test signal and sending the qualified test signal to an image management platform, and generating a color detection signal and sending the color detection signal to a color regulation unit after the image management platform receives the qualified test signal; if the entropy value of the corresponding image is less than the threshold value of the entropy value, judging that the corresponding image test is unqualified, generating a test unqualified signal and sending the test unqualified signal to a mobile phone terminal of an operator; the images are subjected to test evaluation, the image quality is further improved, the storage space occupied by inferior images is reduced, and the working efficiency of image quality management is improved.
Drawings
In order to facilitate understanding for those skilled in the art, the present invention will be further described with reference to the accompanying drawings.
Fig. 1 is a schematic block diagram of the present invention.
Detailed Description
The technical solutions of the present invention will be described clearly and completely with reference to the following embodiments, and it should be understood that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
As shown in fig. 1, the image quality management system based on the neural network includes an image detection unit, an image evaluation unit, a color adjustment unit, a test evaluation unit, a selection storage unit, an image management platform, a registration unit and a database;
the registration login unit is used for submitting management personnel information and operator information through a mobile phone terminal by a manager and an operator, and sending the successfully registered management personnel information and operator information to the database for storage, wherein the management personnel information comprises the name, the age, the time of entry and the mobile phone number of real name authentication of the operator, and the operator information comprises the name, the age, the time of entry and the mobile phone number of real name authentication of the operator;
the image evaluation unit is used for analyzing image raw data so as to evaluate an image, wherein the image raw data comprises a brightness mean value of the image, a pixel gray value standard deviation of the image and an average gradient of the image, the image is marked as i, i is 1, 2, … …, n, n is a positive integer, and the specific analysis and evaluation process is as follows:
the method comprises the steps of firstly, acquiring the number of pixels of an image to be evaluated in the horizontal direction and the number of pixels in the vertical direction in real time, marking the number of pixels of the image in the horizontal direction and the number of pixels in the vertical direction as X and Y respectively, and then marking the resolution of the image to be evaluated as X multiplied by Y;
step two, obtaining any pixel point in the image to be evaluated, marking the pixel point as F (s, e), substituting the pixel point into a calculation formula to obtain the brightness mean value of the image to be evaluated, marking the brightness mean value of the image to be evaluated as LJi, and the calculation formula is as follows:
wherein, alpha is an error correction factor and takes the value of 2.3654512;
step three, substituting the pixel points and the brightness mean value of the image to be evaluated into a calculation formula to obtain the pixel gray value standard deviation of the image to be evaluated, marking the pixel gray value standard deviation of the image to be evaluated as BZi, wherein the calculation formula is as follows:
and substituting the pixel points into a calculation formula to obtain the average gradient of the image to be evaluated, marking the average gradient of the image to be evaluated as TDi, wherein the calculation formula is as follows:
wherein, β is a fixed threshold value set, and the value is 5.63212302;
fourthly, calculating the evaluation coefficient of the brightness mean value of the image, the standard deviation of the pixel gray value of the image and the average gradient of the image, wherein the calculation formula is
Wherein PGi is the evaluation coefficient of the image, a1, a2 and a3 are all proportionality coefficients, and a1 > a2 > a3 > 0;
step five, comparing the evaluation coefficient of the image with an evaluation coefficient threshold value:
if the evaluation coefficient of the image is larger than or equal to the evaluation coefficient threshold value, judging that the corresponding image is qualified for evaluation, generating an evaluation qualified signal and sending the evaluation qualified signal to the image management platform;
if the evaluation coefficient of the image is smaller than the evaluation coefficient threshold value, judging that the corresponding image is unqualified in evaluation, generating an unqualified evaluation signal and sending the unqualified evaluation signal to a mobile phone terminal of an operator, and after receiving the unqualified evaluation signal, the operator performs numerical verification on the brightness mean value of the image, the pixel gray value standard deviation of the image and the average gradient of the image;
the image management platform receives the image evaluation qualified signal and then generates an image detection signal, and sends the image detection signal to the image detection unit, the image detection unit receives the image detection signal and then analyzes the existing data of the image, so as to detect the image, the existing data of the image comprises compressed data, line-pair data and cycle data, the compressed data is the number of characters reduced in the encoding and compressing process of the image data, the line-pair data is the number of black and white lines in the image with equal width per millimeter, the cycle data is the number of image space cycles per millimeter, and the specific analysis and detection process is as follows:
step S1: acquiring the number of characters reduced in the encoding and compressing process of the image data, and marking the number of characters reduced in the encoding and compressing process of the image data as ZFi;
step S2: acquiring black and white line logarithm in the image with the same width per millimeter, and marking the black and white line logarithm in the image with the same width per millimeter as DSi;
step S3: acquiring the number of image space cycles per millimeter, and marking the number of image space cycles per millimeter as ZQi;
step S4: by calculation formula
Acquiring detection coefficients JCi of the image, wherein v1, v2 and v3 are all proportionality coefficients, and v1 is more than v2 is more than v3 is more than 0;
step S5: compare the detection coefficient JCi of the image to a detection coefficient threshold of the image:
if the detection coefficient JCi of the image is not less than the detection coefficient threshold of the image, judging that the corresponding image is qualified for detection, generating an image qualified detection signal and sending the image qualified detection signal to the image management platform, and after receiving the image qualified detection signal, generating an image test signal and sending the image test signal to the test evaluation unit by the image management platform;
if the detection coefficient JCi of the image is smaller than the detection coefficient threshold of the image, judging that the corresponding image is unqualified, generating an unqualified image detection signal and sending the unqualified image detection signal to the mobile phone terminal of the operator;
the test evaluation unit is used for performing test evaluation on the image so as to detect the image, and the specific test detection process is as follows:
step SS 1: randomly selecting ten testers at different age ends, observing the images by the ten testers, counting the content information quantity in the images by the ten testers after the observation is finished, calculating the average content information quantity of the images by averaging, and marking the average content information quantity as PJSi;
step SS 2: then acquiring an image gray distribution set P ═ (P1, P2, … …, Pn) according to the gray distribution of the image, then acquiring the probability of the gray value appearing in the image, and marking the probability of the gray value appearing in the image as P (t);
step SS 3: acquiring an entropy value of a corresponding image through a calculation formula, wherein the calculation formula is as follows:
wherein Si is an entropy value of a corresponding image, and 64 is a fixed gray level threshold value of the image;
step SS 4: comparing the entropy value of the corresponding image with a threshold value of entropy value:
if the entropy value of the corresponding image is larger than or equal to the threshold value of the entropy value, judging that the corresponding image is qualified in the test, generating a qualified test signal and sending the qualified test signal to the image management platform, and after receiving the qualified test signal, the image management platform generates a color detection signal and sends the color detection signal to the color adjusting unit;
if the entropy value of the corresponding image is less than the threshold value of the entropy value, judging that the corresponding image test is unqualified, generating a test unqualified signal and sending the test unqualified signal to a mobile phone terminal of an operator;
the color adjusting unit analyzes the color data of the image after receiving the color detection signal, so as to adjust the color of the image, wherein the color data of the image is the total amount of the colors in the color of the image, the ratio of the number of the colors of the warm color system to the number of the colors of the cold color system, and the saturation of the colors, and the adjusting process is specifically analyzed and adjusted as follows:
step L1: acquiring the total number of colors in the colors of the image, and marking the total number of the colors in the colors of the image as ZSLI;
step L2: acquiring the color quantity ratio of a warm color system and a cold color system in the colors of the image, and marking the color quantity ratio of the warm color system and the cold color system in the colors of the image as SBZi;
step L3: acquiring the color saturation of the image, and marking the color saturation of the image as BGDi;
step L4: by calculation formula
Acquiring color adjustment coefficients TJi of the image, wherein s1, s2 and s3 are all proportionality coefficients, and s1 is more than s2 is more than s3 is more than 0;
step L5: compare the color-toning coefficient TJi for an image to a color-toning coefficient threshold for the image:
if the color adjustment coefficient TJi of the image is not less than the color adjustment coefficient threshold of the image, judging that the corresponding image does not need to be adjusted, generating a color detection qualified signal and sending the color detection qualified signal to the image management platform;
if the color adjustment coefficient TJi of the image is less than the color adjustment coefficient threshold value of the image, judging that the corresponding image needs to be adjusted, generating an unqualified color detection signal and sending the unqualified color detection signal to a mobile phone terminal of a manager;
the image management platform generates an image storage signal and sends the image storage signal to the selective storage unit after receiving the evaluation qualified signal, the image detection qualified signal, the test qualified signal and the color detection qualified signal, and the selective storage unit stores the image after receiving the image storage signal, wherein the specific storage process is as follows:
step T1: setting an independent storage area in a database, and marking the independent storage area as o, o is 1, 2, … …, m is a positive integer;
step T2: acquiring a space memory of an independent storage area, marking the space memory of the independent storage area as NCo, sequencing the independent storage area according to the numerical value of the space memory of the independent storage area, and sending the sequenced independent storage area to an image management platform;
step T3: sorting the images to be stored according to the size of the memory data of the images to be stored, and sending the sorted images to be stored to an image management platform;
step T4: after receiving the sequenced independent storage areas and the sequenced images to be stored, the image management platform matches the independent storage areas and the images to be stored one by one according to the sequence, then generates a protection demand signal and sends the protection demand signal to a mobile phone terminal of a manager, after the manager receives the protection demand signal, if the corresponding image needs to be protected, a determined protection signal is generated and sent to the image management platform, and if the corresponding image does not need to be protected, a signal which does not need to be protected is generated and sent to the image management platform;
step T5: and after receiving the determined protection signal, the image management platform sets a single storage path for the corresponding protection image, and limits the use times of the single storage path, wherein the single storage path represents the only storage path for setting the corresponding protection image.
The image quality management method based on the neural network comprises the following specific steps:
step P1: image evaluation, namely analyzing original image data through an image evaluation unit so as to evaluate the image, and then sending an evaluation qualified signal to an image management platform;
step P2: the image detection is to analyze the existing data of the image after receiving the image detection signal through the image detection unit so as to detect the image, and then to send a detection qualified signal to the image management platform;
step P3: image test evaluation, namely performing test evaluation on the image through a test evaluation unit so as to detect the image, and then sending a test qualified signal to an image management platform;
step P4: the image color adjustment is to analyze the color data of the image after receiving the color detection signal through the color adjustment unit so as to adjust the color of the image, and then the color detection qualified signal is sent to the image management platform;
step P5: and selecting and storing the image, wherein the image management platform generates an image storage signal and sends the image storage signal to the selection storage unit after receiving the evaluation qualified signal, the image detection qualified signal, the test qualified signal and the color detection qualified signal, and then stores the image after receiving the image storage signal through the selection storage unit.
The working principle of the invention is as follows:
when the image quality management system based on the neural network works, analyzing original image data through an image evaluation unit so as to evaluate an image, acquiring the number of pixels of the image to be evaluated in the horizontal direction and the number of pixels of the image to be evaluated in the vertical direction in real time, respectively marking the number of pixels of the image in the horizontal direction and the number of pixels of the image in the vertical direction as X and Y, and then marking the resolution of the image to be evaluated as X multiplied by Y; obtaining any pixel point in an image to be evaluated, marking the pixel point as F (s, e), substituting the pixel point into a calculation formula to obtain a brightness mean value of the image to be evaluated, a pixel gray value standard difference of the image to be evaluated and an average gradient of the image to be evaluated, carrying out evaluation coefficient calculation on the brightness mean value of the image, the pixel gray value standard difference of the image and the average gradient of the image, judging that the corresponding image is qualified for evaluation if an evaluation coefficient of the image is more than or equal to an evaluation coefficient threshold value, generating an evaluation qualified signal and sending the evaluation qualified signal to an image management platform; and if the evaluation coefficient of the image is less than the evaluation coefficient threshold value, judging that the corresponding image is unqualified in evaluation, generating an unqualified evaluation signal and sending the unqualified evaluation signal to a mobile phone terminal of an operator, and after receiving the unqualified evaluation signal, the operator performs numerical verification on the brightness mean value of the image, the pixel gray value standard deviation of the image and the average gradient of the image.
The above calculation formulas are all calculated by taking the numerical value of the dimension-removed data, the formula is a formula of the latest real situation obtained by software simulation of a large amount of collected data, and the preset parameters in the formula are set by the technical personnel in the field according to the actual situation.
The foregoing is merely exemplary and illustrative of the present invention and various modifications, additions and substitutions may be made by those skilled in the art to the specific embodiments described without departing from the scope of the invention as defined in the following claims.
Claims (7)
1. The image quality management system based on the neural network is characterized by comprising an image detection unit, an image evaluation unit, a color adjustment unit, a test evaluation unit, a selection storage unit, an image management platform, a registration unit and a database;
the image evaluation unit is used for analyzing image raw data so as to evaluate an image, wherein the image raw data comprises a brightness mean value of the image, a pixel gray value standard deviation of the image and an average gradient of the image, the image is marked as i, i is 1, 2, … …, n, n is a positive integer, and the specific analysis and evaluation process is as follows:
the method comprises the steps of firstly, acquiring the number of pixels of an image to be evaluated in the horizontal direction and the number of pixels in the vertical direction in real time, marking the number of pixels of the image in the horizontal direction and the number of pixels in the vertical direction as X and Y respectively, and then marking the resolution of the image to be evaluated as X multiplied by Y;
step two, obtaining any pixel point in the image to be evaluated, marking the pixel point as F (s, e), substituting the pixel point into a calculation formula to obtain the brightness mean value of the image to be evaluated, marking the brightness mean value of the image to be evaluated as LJi, and the calculation formula is as follows:
wherein, alpha is an error correction factor and takes the value of 2.3654512;
step three, substituting the pixel points and the brightness mean value of the image to be evaluated into a calculation formula to obtain the pixel gray value standard deviation of the image to be evaluated, marking the pixel gray value standard deviation of the image to be evaluated as BZi, wherein the calculation formula is as follows:
and substituting the pixel points into a calculation formula to obtain the average gradient of the image to be evaluated, marking the average gradient of the image to be evaluated as TDi, wherein the calculation formula is as follows:
wherein, β is a fixed threshold value set, and the value is 5.63212302;
fourthly, calculating the evaluation coefficient of the brightness mean value of the image, the standard deviation of the pixel gray value of the image and the average gradient of the image, wherein the calculation formula is
Wherein PGi is the evaluation coefficient of the image, a1, a2 and a3 are all proportionality coefficients, and a1 > a2 > a3 > 0;
step five, comparing the evaluation coefficient of the image with an evaluation coefficient threshold value:
if the evaluation coefficient of the image is larger than or equal to the evaluation coefficient threshold value, judging that the corresponding image is qualified for evaluation, generating an evaluation qualified signal and sending the evaluation qualified signal to the image management platform;
and if the evaluation coefficient of the image is less than the evaluation coefficient threshold value, judging that the corresponding image is unqualified in evaluation, generating an unqualified evaluation signal and sending the unqualified evaluation signal to a mobile phone terminal of an operator, and after receiving the unqualified evaluation signal, the operator performs numerical verification on the brightness mean value of the image, the pixel gray value standard deviation of the image and the average gradient of the image.
2. The image quality management system based on the neural network as claimed in claim 1, wherein the image management platform receives the image evaluation qualified signal to generate an image detection signal, and sends the image detection signal to the image detection unit, the image detection unit receives the image detection signal and then analyzes the existing image data to detect the image, the existing image data includes compressed data, line-pair data and period data, the compressed data is the number of characters reduced in the encoding and compressing process of the image data, the line-pair data is the number of black and white line-pairs in the image with the same width per millimeter, and the period data is the number of image space cycles per millimeter, and the specific analysis and detection process is as follows:
step S1: acquiring the number of characters reduced in the encoding and compressing process of the image data, and marking the number of characters reduced in the encoding and compressing process of the image data as ZFi;
step S2: acquiring black and white line logarithm in the image with the same width per millimeter, and marking the black and white line logarithm in the image with the same width per millimeter as DSi;
step S3: acquiring the number of image space cycles per millimeter, and marking the number of image space cycles per millimeter as ZQi;
step S4: by calculation formula
Acquiring detection coefficients JCi of the image, wherein v1, v2 and v3 are all proportionality coefficients, and v1 is more than v2 is more than v3 is more than 0;
step S5: compare the detection coefficient JCi of the image to a detection coefficient threshold of the image:
if the detection coefficient JCi of the image is not less than the detection coefficient threshold of the image, judging that the corresponding image is qualified for detection, generating an image qualified detection signal and sending the image qualified detection signal to an image management platform, and after receiving the image qualified detection signal, the image management platform generates an image test signal and sends the image test signal to a test evaluation unit;
and if the detection coefficient JCi of the image is less than the detection coefficient threshold of the image, judging that the corresponding image is unqualified, generating an unqualified image detection signal and sending the unqualified image detection signal to the mobile phone terminal of the operator.
3. The image quality management system based on the neural network as claimed in claim 1, wherein the test evaluation unit is configured to perform a test evaluation on the image so as to detect the image, and the specific test detection process is as follows:
step SS 1: randomly selecting ten testers at different age ends, observing the images by the ten testers, counting the content information quantity in the images by the ten testers after the observation is finished, calculating the average content information quantity of the images by averaging, and marking the average content information quantity as PJSi;
step SS 2: then acquiring an image gray distribution set P ═ (P1, P2, … …, Pn) according to the gray distribution of the image, then acquiring the probability of the gray value appearing in the image, and marking the probability of the gray value appearing in the image as P (t);
step SS 3: acquiring an entropy value of a corresponding image through a calculation formula, wherein the calculation formula is as follows:
wherein Si is an entropy value of a corresponding image, and 64 is a fixed gray level threshold value of the image;
step SS 4: comparing the entropy value of the corresponding image with a threshold value of entropy value:
if the entropy value of the corresponding image is larger than or equal to the threshold value of the entropy value, judging that the corresponding image is qualified in the test, generating a qualified test signal and sending the qualified test signal to the image management platform, and after receiving the qualified test signal, the image management platform generates a color detection signal and sends the color detection signal to the color adjusting unit;
and if the entropy value of the corresponding image is less than the threshold value of the entropy value, judging that the corresponding image test is unqualified, generating a test unqualified signal and sending the test unqualified signal to the mobile phone terminal of the operator.
4. The image quality management system based on the neural network as claimed in claim 1, wherein the color adjusting unit analyzes the color data of the image after receiving the color detection signal, so as to adjust the color of the image, and the color data of the image is the total number of colors in the color of the image, the ratio of the number of colors in the warm color system and the cold color system, and the saturation of the color, and the specific analyzing and adjusting process is as follows:
step L1: acquiring the total number of colors in the colors of the image, and marking the total number of the colors in the colors of the image as ZSLI;
step L2: acquiring the color quantity ratio of a warm color system and a cold color system in the colors of the image, and marking the color quantity ratio of the warm color system and the cold color system in the colors of the image as SBZi;
step L3: acquiring the color saturation of the image, and marking the color saturation of the image as BGDi;
step L4: by calculation formula
Acquiring color adjustment coefficients TJi of the image, wherein s1, s2 and s3 are all proportionality coefficients, and s1 is more than s2 is more than s3 is more than 0;
step L5: compare the color-toning coefficient TJi for an image to a color-toning coefficient threshold for the image:
if the color adjustment coefficient TJi of the image is not less than the color adjustment coefficient threshold of the image, judging that the corresponding image does not need to be adjusted, generating a color detection qualified signal and sending the color detection qualified signal to the image management platform;
and if the color adjustment coefficient TJi of the image is less than the color adjustment coefficient threshold value of the image, judging that the corresponding image needs to be adjusted, generating a unqualified color detection signal and sending the unqualified color detection signal to the mobile phone terminal of the administrator.
5. The image quality management system based on the neural network as claimed in claim 1, wherein the image management platform generates an image saving signal and sends the image saving signal to the selective saving unit after receiving the evaluation qualified signal, the image detection qualified signal, the test qualified signal and the color detection qualified signal, and the selective saving unit stores the image after receiving the image saving signal, wherein the specific storage process is as follows:
step T1: setting an independent storage area in a database, and marking the independent storage area as o, o is 1, 2, … …, m is a positive integer;
step T2: acquiring a space memory of an independent storage area, marking the space memory of the independent storage area as NCo, sequencing the independent storage area according to the numerical value of the space memory of the independent storage area, and sending the sequenced independent storage area to an image management platform;
step T3: sorting the images to be stored according to the size of the memory data of the images to be stored, and sending the sorted images to be stored to an image management platform;
step T4: after receiving the sequenced independent storage areas and the sequenced images to be stored, the image management platform matches the independent storage areas and the images to be stored one by one according to the sequence, then generates a protection demand signal and sends the protection demand signal to a mobile phone terminal of a manager, after the manager receives the protection demand signal, if the corresponding image needs to be protected, a determined protection signal is generated and sent to the image management platform, and if the corresponding image does not need to be protected, a signal which does not need to be protected is generated and sent to the image management platform;
step T5: and after receiving the determined protection signal, the image management platform sets a single storage path for the corresponding protection image, and limits the use times of the single storage path, wherein the single storage path represents the only storage path for setting the corresponding protection image.
6. The image quality management system according to claim 1, wherein the registration login unit is configured to allow a manager and an operator to submit manager information and operator information via a mobile phone terminal, and to send the manager information and the operator information that are successfully registered to the database for storage, the manager information includes a name, an age, an attendance time of the manager and a mobile phone number for personal real name authentication, and the operator information includes a name, an age, an attendance time of the operator and a mobile phone number for personal real name authentication.
7. The image quality management method based on the neural network is characterized by comprising the following specific management method steps:
step P1: image evaluation, namely analyzing original image data through an image evaluation unit so as to evaluate the image, and then sending an evaluation qualified signal to an image management platform;
step P2: the image detection is to analyze the existing data of the image after receiving the image detection signal through the image detection unit so as to detect the image, and then to send a detection qualified signal to the image management platform;
step P3: image test evaluation, namely performing test evaluation on the image through a test evaluation unit so as to detect the image, and then sending a test qualified signal to an image management platform;
step P4: the image color adjustment is to analyze the color data of the image after receiving the color detection signal through the color adjustment unit so as to adjust the color of the image, and then the color detection qualified signal is sent to the image management platform;
step P5: and selecting and storing the image, wherein the image management platform generates an image storage signal and sends the image storage signal to the selection storage unit after receiving the evaluation qualified signal, the image detection qualified signal, the test qualified signal and the color detection qualified signal, and then stores the image after receiving the image storage signal through the selection storage unit.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110024819.1A CN112767327B (en) | 2021-01-08 | 2021-01-08 | Image quality management system and method based on neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110024819.1A CN112767327B (en) | 2021-01-08 | 2021-01-08 | Image quality management system and method based on neural network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112767327A true CN112767327A (en) | 2021-05-07 |
| CN112767327B CN112767327B (en) | 2021-09-07 |
Family
ID=75701076
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110024819.1A Active CN112767327B (en) | 2021-01-08 | 2021-01-08 | Image quality management system and method based on neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112767327B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114025104A (en) * | 2021-11-03 | 2022-02-08 | 江苏金视传奇科技有限公司 | Color image display device and color image display system |
| CN114510989A (en) * | 2021-12-23 | 2022-05-17 | 中国科学院软件研究所 | Normative evaluation method, device and equipment for image data set |
| CN116503365A (en) * | 2023-05-04 | 2023-07-28 | 无锡化工装备股份有限公司 | Machine vision-based three-dimensional condensing finned tube rolling real-time monitoring method |
| CN117238229A (en) * | 2023-11-16 | 2023-12-15 | 深圳市诚电科技有限公司 | Automatic alignment method and system based on stereoscopic liquid crystal display |
| CN117611578A (en) * | 2024-01-17 | 2024-02-27 | 深圳市新良田科技股份有限公司 | Image processing method and image processing system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2007110606A (en) * | 2005-10-17 | 2007-04-26 | Kddi Corp | Allocating image forming method, its program and secret image decoding method |
| CN101216881A (en) * | 2007-12-28 | 2008-07-09 | 北京中星微电子有限公司 | A method and device for automatic image acquisition |
| CN104408707A (en) * | 2014-10-28 | 2015-03-11 | 哈尔滨工业大学 | Rapid digital imaging fuzzy identification and restored image quality assessment method |
| CN107274403A (en) * | 2017-06-30 | 2017-10-20 | 长安大学 | A kind of evaluation method of flotation surface quality |
-
2021
- 2021-01-08 CN CN202110024819.1A patent/CN112767327B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2007110606A (en) * | 2005-10-17 | 2007-04-26 | Kddi Corp | Allocating image forming method, its program and secret image decoding method |
| CN101216881A (en) * | 2007-12-28 | 2008-07-09 | 北京中星微电子有限公司 | A method and device for automatic image acquisition |
| CN104408707A (en) * | 2014-10-28 | 2015-03-11 | 哈尔滨工业大学 | Rapid digital imaging fuzzy identification and restored image quality assessment method |
| CN107274403A (en) * | 2017-06-30 | 2017-10-20 | 长安大学 | A kind of evaluation method of flotation surface quality |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114025104A (en) * | 2021-11-03 | 2022-02-08 | 江苏金视传奇科技有限公司 | Color image display device and color image display system |
| CN114025104B (en) * | 2021-11-03 | 2024-02-20 | 江苏金视传奇科技有限公司 | Color image display device and color image display system |
| CN114510989A (en) * | 2021-12-23 | 2022-05-17 | 中国科学院软件研究所 | Normative evaluation method, device and equipment for image data set |
| CN114510989B (en) * | 2021-12-23 | 2022-10-25 | 中国科学院软件研究所 | Normative evaluation method, device and equipment for image data set |
| CN116503365A (en) * | 2023-05-04 | 2023-07-28 | 无锡化工装备股份有限公司 | Machine vision-based three-dimensional condensing finned tube rolling real-time monitoring method |
| CN117238229A (en) * | 2023-11-16 | 2023-12-15 | 深圳市诚电科技有限公司 | Automatic alignment method and system based on stereoscopic liquid crystal display |
| CN117238229B (en) * | 2023-11-16 | 2024-02-09 | 深圳市诚电科技有限公司 | Automatic alignment method and system based on stereoscopic liquid crystal display |
| CN117611578A (en) * | 2024-01-17 | 2024-02-27 | 深圳市新良田科技股份有限公司 | Image processing method and image processing system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112767327B (en) | 2021-09-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112767327B (en) | Image quality management system and method based on neural network | |
| Gu et al. | Learning a no-reference quality assessment model of enhanced images with big data | |
| Yue et al. | Effective and efficient blind quality evaluator for contrast distorted images | |
| Ma et al. | Reduced-reference image quality assessment in reorganized DCT domain | |
| CN101562675B (en) | No-reference image quality evaluation method based on Contourlet transform | |
| US20240005468A1 (en) | Image distortion evaluation method and apparatus, and computer device | |
| Liu et al. | Pre-attention and spatial dependency driven no-reference image quality assessment | |
| CN109191428A (en) | Full-reference image quality evaluating method based on masking textural characteristics | |
| CN113590906A (en) | Land utilization change proof checking system | |
| CN108052918A (en) | A kind of person's handwriting Compare System and method | |
| Zhang et al. | Perceptual video coding with block-level staircase just noticeable distortion | |
| CN114462646A (en) | Pole number plate identification method and system based on contact network safety inspection | |
| CN109409305A (en) | A kind of facial image clarity evaluation method and device | |
| CN116311088B (en) | Construction safety monitoring method based on construction site | |
| CN107945154A (en) | Color image quality evaluation method based on quaternary number discrete cosine transform | |
| CN115550658B (en) | Data transmission method based on intelligent campus management platform | |
| CN110135274B (en) | Face recognition-based people flow statistics method | |
| Yang et al. | EHNQ: Subjective and objective quality evaluation of enhanced night-time images | |
| CN115146126A (en) | Artwork display recommendation system based on intelligent cloud platform | |
| CN114596287A (en) | Image quality determination method, device, equipment and storage medium | |
| Zhu et al. | Siqd: Surveillance image quality database and performance evaluation for objective algorithms | |
| CN115049426A (en) | Artificial intelligence algorithm model for personalized skin care recommendation | |
| Zhai et al. | Fuzzy comprehensive evaluation method and its application in subjective quality assessment for compressed remote sensing images | |
| CN107610101B (en) | Method for measuring visual balance quality of digital image | |
| Kung et al. | Investigation of the image quality assessment using neural networks and structure similarty |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |