CN113362354A - Method, system, terminal and storage medium for evaluating quality of tone mapping image - Google Patents
Method, system, terminal and storage medium for evaluating quality of tone mapping image Download PDFInfo
- Publication number
- CN113362354A CN113362354A CN202110494547.1A CN202110494547A CN113362354A CN 113362354 A CN113362354 A CN 113362354A CN 202110494547 A CN202110494547 A CN 202110494547A CN 113362354 A CN113362354 A CN 113362354A
- Authority
- CN
- China
- Prior art keywords
- local
- exposure
- image
- tone mapping
- tone
- 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
- 238000013507 mapping Methods 0.000 title claims abstract description 68
- 238000000034 method Methods 0.000 title claims abstract description 49
- 238000000605 extraction Methods 0.000 claims abstract description 47
- 238000013441 quality evaluation Methods 0.000 claims abstract description 32
- 230000002159 abnormal effect Effects 0.000 claims abstract description 21
- 238000012549 training Methods 0.000 claims abstract description 20
- 238000000638 solvent extraction Methods 0.000 claims abstract description 17
- 238000012360 testing method Methods 0.000 claims abstract description 16
- 238000011156 evaluation Methods 0.000 claims abstract description 7
- 239000013598 vector Substances 0.000 claims description 27
- 238000005192 partition Methods 0.000 claims description 11
- 239000011159 matrix material Substances 0.000 claims description 8
- 238000004590 computer program Methods 0.000 claims description 6
- 238000005259 measurement Methods 0.000 claims description 5
- 238000004364 calculation method Methods 0.000 claims description 4
- 230000008859 change Effects 0.000 claims description 3
- 230000002093 peripheral effect Effects 0.000 claims description 3
- 238000005516 engineering process Methods 0.000 abstract description 5
- 238000004422 calculation algorithm Methods 0.000 description 15
- 230000006870 function Effects 0.000 description 4
- 230000008569 process Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 238000002474 experimental method Methods 0.000 description 2
- 230000004927 fusion Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012805 post-processing Methods 0.000 description 2
- 238000003384 imaging method Methods 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000001303 quality assessment method Methods 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
- 238000012800 visualization Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
- G06N20/10—Machine learning using kernel methods, e.g. support vector machines [SVM]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/187—Segmentation; Edge detection involving region growing; involving region merging; involving connected component labelling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/40—Analysis of texture
- G06T7/41—Analysis of texture based on statistical description of texture
- G06T7/45—Analysis of texture based on statistical description of texture using co-occurrence matrix computation
-
- 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
- G06T2207/10—Image acquisition modality
- G06T2207/10024—Color image
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- 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
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Probability & Statistics with Applications (AREA)
- Artificial Intelligence (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- General Engineering & Computer Science (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a quality evaluation method, a system, a terminal and a storage medium of a tone mapping image, wherein the method comprises the following steps: dividing the tone mapping image into a normal exposure area and an abnormal exposure area according to the exposure value; performing regional local feature extraction on the normal exposure region and the abnormal exposure region; extracting local structural features; extracting surface structure features; extracting edge structure features; and (5) training and testing the model. The system comprises: the device comprises a partitioning unit, a partitioning local feature extraction unit, a local structure feature extraction unit, a surface structure feature extraction unit, an edge structure feature extraction unit, a model training unit and a test unit. By the method and the device, the performance and subjective consistency of the non-reference tone mapping image quality evaluation technology are improved, and the evaluation efficiency is improved.
Description
Technical Field
The present invention relates to the field of image processing and evaluation technologies, and in particular, to a method, a system, a terminal, and a storage medium for evaluating quality of a tone-mapped image.
Background
High Dynamic Range (HDR) imaging techniques can vividly describe real scenes of greater brightness variation from sunlight directly to faint starlight. However, most displays are now standard 8-bit displays, and cannot reproduce the wide range of luminance variations of HDR images, resulting in some loss of important visual information. To address this issue, more and more Tone Mapping Operators (TMOs) convert HDR images into Standard Dynamic Range (SDR) images for the visualization of HDR images on standard 8-bit displays. Information loss is inevitable during tone mapping, but tone-mapped images (TMIs) created from suitable tone-mapped images (TMOs) can provide better quality perception of brightness, contrast and detail than normal SDR images. However, due to the diversity of image structures and contents, there is currently no universal TMO that can be effectively applied to various HDR images. Therefore, Image Quality Assessment (IQA) is often used to predict the quality of tone-mapped images (TMIs) in order to optimize the tone mapping process. Compared with a subjective picture quality evaluation method carried out manually, the objective hue quality evaluation index without manual participation can be used for automatically selecting the hue quality evaluation index with the best quality from candidate hue quality evaluation indexes generated by different hue quality evaluation indexes on line, and can also be used for selecting the optimal parameter in hue mapping.
Since the original HDR image is difficult to obtain in practical applications, the no-reference tone mapping image quality evaluation algorithm has been a major concern. Therefore, it is urgently needed to provide a quality evaluation method and system for tone-mapped images with better performance.
Disclosure of Invention
The invention provides a tone mapping image quality evaluation method, a system, a terminal and a storage medium based on exposure attribute partition aiming at the problems in the prior art, so that the performance and subjective consistency of the quality evaluation technology of the non-reference Tone Mapping Image (TMI) are improved, and the evaluation efficiency is improved.
In order to solve the technical problems, the invention is realized by the following technical scheme:
the invention provides a quality evaluation method of tone mapping image partitioned based on exposure attribute, which comprises the following steps:
s11: partitioning the tone-mapped image based on the exposure attributes; dividing the tone mapping image into a normal exposure area and an abnormal exposure area according to the exposure value;
s12: performing regional local feature extraction on the normal exposure region and the abnormal exposure region;
s13: extracting local structural features of the tone mapping image, wherein the local structural features are local texture features;
s14: extracting surface structure features of the tone mapping image, wherein the surface structure features are texture features;
s15: extracting edge structure characteristics of the tone mapping image, wherein the edge structure characteristics measure the brightness of the edges of the tone mapping image and the divergence degree of edge areas;
s16: combining the features extracted from S12-S15 into an integral feature vector, taking the corresponding subjective quality score provided in an image database as a label, training a support vector regressor through continuous iteration, and evaluating by adopting the trained support vector regressor to obtain the predicted objective quality score of the image.
Preferably, the S11 further includes:
s21: converting the RGB space of the tone mapping image into an LAB space, and calculating the exposure of a pixel level based on the value of an L channel;
s22: the tone-mapped image is divided into a normal exposure area and an abnormal exposure area according to an exposure value using a one-dimensional maximum entropy method.
Preferably, the performing of the exposure measurement at the pixel level based on the value of the L channel in S21 further comprises:
s211: normalizing the L-channel brightness value to an interval [0,1 ];
s212: obtaining an exposure value L for each pixel i using a Gaussian weighting functioni：
Li＝(exp(-(i-it)2/2σ2))；
Wherein itTo be a luminance threshold, σ is a width parameter of the function.
Preferably, the S12 further includes:
s41: measuring the contrast condition of the local small blocks by using the color index aiming at the normal exposure area, and generating local color feature vectors;
s42: for an abnormal exposure area, firstly, calculating detail information quantity by using an information entropy method, and then measuring local gradient change by using the local contrast of the d-th pixel obtained by the following formula;
wherein idIs the gray value of the d-th pixel point, itThe gray value of the p-th neighbor pixel point in the peripheral local area of the d-th pixel point is shown, and p is the number of the local neighbor pixel points. .
Preferably, the S13 further includes: performing local texture measurement and calculation by using a centrosymmetric local binary pattern operator, and generating 16 local structure feature vectors by each tone mapping image; wherein the centrosymmetric local binary pattern operator CS-LBP is:
μ(a,b)＝a*21+b*20；
where N is the total number of neighbors of the pixel involved, mcGray value of the central pixel, mi(i-0, 1, …, (N/2) -1) is the gray scale value of the adjacent pixel.
Preferably, the S14 further includes: converting the L channel of the tone mapping image into a gray level co-occurrence matrix, and extracting three characteristics which embody the thin edge characteristics: energy E, correlation C, and uniformity V; wherein,
μ and σ denote the mean and variance of the pixel, where
Wherein G (x, y) is the x-th row and y-th column of the gray level co-occurrence matrix, and k is the scale range of the gray level domain.
Preferably, the S15 further includes:
s71: extracting the edge of each tone mapping image by using a Canny operator;
s72: calculating the average value H of the brightness of small blocks with preset sizes around the edge pixel points extracted in the step S711Variance H2Degree of sum deviation H3To measure the brightness of the image edge and the degree of divergence of the edge region; wherein,
H1＝E(he)；
H2＝E(he 2)-E(he)2；
preferably, after S162, the method further includes:
s163: the objective quality scores of the images are compared with corresponding subjective quality scores provided in an image database to verify prediction accuracy.
The invention also provides a quality evaluation system of tone mapping image based on exposure attribute partition, which is used for realizing the quality evaluation method of tone mapping image based on exposure attribute partition, and comprises the following steps: the device comprises a partitioning unit, a partitioning local feature extraction unit, a local structure feature extraction unit, a surface structure feature extraction unit, an edge structure feature extraction unit, a model training unit and a test unit; wherein,
the partitioning unit is used for partitioning the tone mapping image into a normal exposure area and an abnormal exposure area according to an exposure value;
the regional local feature extraction unit is used for dividing the tone mapping image into a normal exposure region and an abnormal exposure region according to an exposure value;
the local structure feature extraction unit is used for extracting local texture features of the tone mapping image;
the surface structure feature extraction unit is used for extracting texture features of the tone mapping image;
the edge structure characteristic extraction unit is used for measuring the brightness of the tone mapping image edge and the divergence degree of the edge area;
the model training unit is used for combining the features extracted by the partition local feature extraction unit, the local structure feature extraction unit, the surface structure feature extraction unit and the edge structure feature extraction unit into an integral feature vector, and training a support vector regressor by continuous iteration by taking corresponding subjective quality scores provided in an image database as labels;
and the test unit is used for inputting the characteristic value of the test data in the data set into the support vector regression trained by the model training unit to obtain the predicted objective quality score of the image.
The invention also provides a quality evaluation terminal of the tone mapping image, which comprises a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein the processor can be used for executing the quality evaluation method of the tone mapping image when executing the program.
The present invention also provides a computer-readable storage medium having stored thereon a computer program for executing the above-mentioned method of evaluating the quality of a tone-mapped image when the program is executed by a processor.
Compared with the prior art, the embodiment of the invention has at least one of the following advantages:
(1) according to the tone mapping image quality evaluation method and system based on exposure attribute partition, the quality of the tone mapping image can be accurately and effectively predicted through feature extraction of different areas and extraction of structural features and edge features;
(2) according to the tone mapping image quality evaluation method and system based on the exposure attribute partition, the structural features are extracted through the central symmetry local binary pattern and the gray level co-occurrence matrix, the edge features are extracted in the edge texture generated based on the Canny operator, and the performance and subjective consistency of the quality evaluation technology of the non-reference Tone Mapping Image (TMI) are further improved.
(3) The quality evaluation method and the quality evaluation system of the tone mapping image based on the exposure attribute partition accurately measure the correlation of local spatial distribution of different exposure areas and the difference of regional color distortion, and improve the quality evaluation efficiency of the tone mapping image.
Drawings
Embodiments of the invention are further described below with reference to the accompanying drawings:
FIG. 1 is a flow chart of a method for evaluating the quality of a tone-mapped image partitioned based on exposure attributes according to an embodiment of the present invention;
FIG. 2 is a diagram of an exposure zone according to one embodiment of the present invention;
fig. 3 is a diagram of edge structure features extracted by Canny operators according to an embodiment of the present invention.
Detailed Description
The present invention will be described in detail with reference to specific examples. The following examples will assist those skilled in the art in further understanding the invention, but are not intended to limit the invention in any way. It should be noted that variations and modifications can be made by persons skilled in the art without departing from the spirit of the invention. All falling within the scope of the present invention.
FIG. 1 is a flowchart of a method for evaluating the quality of a tone-mapped image partitioned based on exposure attributes according to an embodiment of the present invention.
Referring to fig. 1, the method for evaluating the quality of a tone-mapped image partitioned based on exposure attributes of the present embodiment includes:
s11: partitioning the tone-mapped image based on the exposure attributes;
dividing the tone mapping image into a normal exposure area and an abnormal exposure area according to the exposure value;
s12: performing regional local feature extraction on the normal exposure region and the abnormal exposure region;
s13: local structural feature extraction: extracting local texture features of the tone mapping image;
s14: extracting surface structure features: extracting texture features of the tone mapping image;
s15: extracting edge structure characteristics, and measuring the brightness of the tone mapping image edge and the divergence degree of an edge area;
s16: combining the features extracted from S12-S15 into an integral feature vector, taking the corresponding subjective quality score provided in an image database as a label, training a support vector regressor through continuous iteration, and evaluating by adopting the trained support vector regressor to obtain the predicted objective quality score of the image.
In the above embodiment, the steps S12 to S15 do not need to be in the exact order described above, and these steps may be performed in any order.
The embodiment of the invention can accurately and effectively predict the quality of the tone mapping image by extracting the characteristics of different areas and extracting the structural characteristics and the edge characteristics.
In a preferred embodiment, S11 further includes:
s21: converting the RGB space of the tone mapping image into an LAB space, and calculating the exposure of a pixel level based on the value of an L channel;
s22: the tone-mapped image is divided into a normal exposure region and an abnormal exposure region according to the exposure value using the one-dimensional maximum entropy method, as shown in fig. 2, white represents the abnormal exposure region, and black represents the normal exposure region.
In a preferred embodiment, the pixel-level exposure measurement based on the value of the L channel in S21 further comprises:
s211: normalizing the L-channel brightness value to an interval [0,1 ];
s212: obtaining an exposure value L for each pixel i using a Gaussian weighting functioni：
Li＝(exp(-(i-it)2/2σ2))；
Wherein itσ is set to 0.2 as the luminance threshold value.
In a preferred embodiment, S12 further includes:
s41: measuring the contrast condition of the local small blocks by using the color index aiming at the normal exposure area, and generating local color feature vectors;
s42: for an abnormal exposure area, firstly, calculating detail information quantity by using an information entropy method, and then measuring local gradient change by using local contrast obtained by the following formula;
wherein idIs the gray value of the d-th pixel point, itThe gray value of the p-th neighbor pixel point in the peripheral local area of the d-th pixel point is shown, and p is the number of the local neighbor pixel points.
In a preferred embodiment, S13 further includes: performing local texture measurement and calculation by using a centrosymmetric local binary pattern operator, and generating 16 local structure feature vectors by each tone mapping image; wherein the centrosymmetric local binary pattern operator CS-LBP is:
μ(a,b)＝a*21+b*20；
where N is the total number of neighbors of the pixel involved, mcGray value of the central pixel, mi(i-0, 1, …, (N/2) -1) is the gray scale value of the adjacent pixel.
In a preferred embodiment, S14 further includes: converting an L channel of the tone mapping image into a gray level co-occurrence matrix, and extracting three characteristics which can better reflect the characteristics of thin edges: energy E, correlation C, and uniformity V; wherein,
where μ and σ represent the mean and variance of the pixel,
g (x, y) is the x-th row and y-th column of the gray level co-occurrence matrix, and k is the scale range of the gray level domain.
In the embodiment, the structural features are extracted through the central symmetry local binary pattern and the gray level co-occurrence matrix, the edge features are extracted from the edge texture generated based on the Canny operator, and the performance and subjective consistency of the quality evaluation technology of the non-reference Tone Mapping Image (TMI) are further improved.
In a preferred embodiment, S15 further includes:
s71: extracting the edge of each tone mapping image by using a Canny operator;
s72: calculating the mean value H of the 8 multiplied by 8 small block brightness around the edge pixel point extracted by S711Variance H2Degree of sum deviation H3To measure the brightness of the image edge and the degree of divergence of the edge region; wherein,
H1＝E(he)；
H2＝E(he 2)-E(he)2；
in the preferred embodiment, the features extracted from S12-S15 are used to train the model and tested, namely:
s161: training a model: combining the features extracted from S12-S15 into an integral feature vector, taking the corresponding subjective quality score provided in an image database as a label, and training a support vector regression through continuous iteration;
s162: and (3) testing: and inputting the characteristic value of the test data in the data set into a trained support vector regression to obtain the predicted objective quality score of the image.
S162 may be followed by:
s163: the image objective quality scores are compared to the corresponding subjective quality scores provided in the image database to verify prediction accuracy.
In another embodiment of the present invention, there is further provided an exposure attribute partition-based tone-mapped image quality evaluation system for implementing the exposure attribute partition-based tone-mapped image quality evaluation method of the above embodiments, including: the device comprises a partitioning unit, a partitioning local feature extraction unit, a local structure feature extraction unit, a surface structure feature extraction unit, an edge structure feature extraction unit, a model training unit and a test unit; wherein the partitioning unit is used for partitioning the tone mapping image into a normal exposure area and an abnormal exposure area according to the exposure value; the regional local feature extraction unit is used for dividing the tone mapping image into a normal exposure region and an abnormal exposure region according to the exposure value; the local structure feature extraction unit is used for extracting local texture features of the tone mapping image; the surface structure feature extraction unit is used for extracting texture features of the tone mapping image; the edge structure characteristic extraction unit is used for measuring the brightness of the tone mapping image edge and the divergence degree of the edge area; the model training unit is used for combining the features extracted by the partition local feature extraction unit, the local structure feature extraction unit, the surface structure feature extraction unit and the edge structure feature extraction unit into an integral feature vector, and training a support vector regressor by continuous iteration by taking the corresponding subjective quality scores provided in an image database as labels; and the test unit is used for inputting the characteristic value of the test data in the data set into the well-trained support vector regression unit of the model training unit to obtain the predicted objective quality score of the image.
Of course, on the basis of the above embodiment, the system may further include: and the verification unit is used for comparing the objective quality scores of the images obtained by the testing unit with the corresponding subjective quality scores provided in the image database to verify the prediction accuracy.
In another embodiment of the present invention, there is also provided a quality evaluation terminal for tone-mapped images, including a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor is operable to execute the quality evaluation method for tone-mapped images.
In another embodiment of the present invention, there is also provided a computer-readable storage medium having stored thereon a computer program for executing the above-described method of evaluating the quality of a tone-mapped image when the program is executed by a processor.
To verify the performance of the present invention, the following was performed in the ESPL-LIVE HDR database. This database is the largest TMI database released at present, with 1811 tone-mapped HDR images generated by tone mapping, multi-exposure fusion algorithms and post-processing methods. In the experiment, a plurality of excellent natural images and TMI quality evaluation algorithms are selected as comparison algorithms, the natural image quality evaluation algorithm is italicized, and three common evaluation indexes are used for measuring the performance of the algorithm, wherein the three common evaluation indexes are respectively as follows: pearson Linear Correlation Coefficient (PLCC), Spireman Rank Correlation Coefficient (SRCC), and mean square error (RMSE). Wherein, the bigger the values of PLCC and SROCC are, the smaller the value of RMSE is, which indicates that the objective image quality evaluation algorithm is better.
In order to improve the accuracy of the method, the experiment adopts that the training and testing process is executed 1000 times on a support vector machine, and finally the average value of the 1000 times of results is taken to represent the performance result of the algorithm. Table 1 shows the overall performance of the present invention and other excellent algorithms in the ESPL-LIVE HDR database. It can be seen that the overall performance of the method of the invention is significantly better than the performance of several other algorithms.
TABLE 1 comparison of the overall performance of the present invention on the ESPL-LIVE HDR database with several mainstream reference-free algorithms
In addition, performance was evaluated for different distortion types (see table 2). The results of the best process performance are indicated in bold. It can be seen that the algorithm provided by the invention has the best performance in tone mapping and post-processing, the third performance in multi-exposure fusion distortion, and the comprehensive performance superior to other algorithms, so that the algorithm of the invention can evaluate the quality of TMI more accurately and effectively.
Table 2 performance evaluation of different distortion types by the present invention and other non-reference quality evaluation algorithms
It should be noted that, the steps in the method provided by the present invention may be implemented by using corresponding modules, devices, units, and the like in the system, and those skilled in the art may refer to the technical solution of the system to implement the step flow of the method, that is, the embodiment in the system may be understood as a preferred example for implementing the method, and details are not described herein.
Those skilled in the art will appreciate that, in addition to implementing the system and its various devices provided by the present invention in purely computer readable program code means, the method steps can be fully programmed to implement the same functions by implementing the system and its various devices in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers, embedded microcontrollers and the like. Therefore, the system and various devices thereof provided by the present invention can be regarded as a hardware component, and the devices included in the system and various devices thereof for realizing various functions can also be regarded as structures in the hardware component; means for performing the functions may also be regarded as structures within both software modules and hardware components for performing the methods.
The embodiments were chosen and described in order to best explain the principles of the invention and the practical application, and not to limit the invention. Any modifications and variations within the scope of the description, which may occur to those skilled in the art, are intended to be within the scope of the invention.
Claims (10)
1. A method for evaluating the quality of a tone-mapped image partitioned based on exposure attributes, comprising:
s11: partitioning the tone-mapped image based on the exposure attributes; dividing the tone mapping image into a normal exposure area and an abnormal exposure area according to the exposure value;
s12: performing regional local feature extraction on the normal exposure region and the abnormal exposure region;
s13: extracting local structural features of the tone mapping image, wherein the local structural features are local texture features;
s14: extracting surface structure features of the tone mapping image, wherein the surface structure features are texture features;
s15: extracting edge structure characteristics of the tone mapping image, wherein the edge structure characteristics measure the brightness of the edges of the tone mapping image and the divergence degree of edge areas;
s16: combining the features extracted from S12-S15 into an integral feature vector, training a support vector regression with the corresponding subjective quality score provided in the image database as a label, and performing test evaluation by using the trained support vector regression to obtain the predicted objective image quality score.
2. The method for evaluating the quality of a tone-mapped image partitioned based on exposure properties according to claim 1, wherein said S11 further comprises:
s21: converting the RGB space of the tone mapping image into an LAB space, and calculating the exposure of a pixel level based on the value of an L channel;
s22: the tone-mapped image is divided into a normal exposure area and an abnormal exposure area according to an exposure value using a one-dimensional maximum entropy method.
3. The method for evaluating the quality of a tone-mapped image partitioned based on exposure properties according to claim 2, wherein said performing pixel-level exposure calculation based on the value of L channel in S21 further comprises:
s211: normalizing the L-channel brightness value to an interval [0,1 ];
s212: obtaining an exposure value L for each pixel i using a Gaussian weighting functioni：
Li＝(exp(-(i-it)2/2σ2))；
Wherein itσ is set to 0.2 as the luminance threshold.
4. The method for evaluating the quality of a tone-mapped image partitioned based on exposure properties according to claim 1, wherein said S12 further comprises:
s41: measuring the contrast condition of the local small blocks by using the color index aiming at the normal exposure area, and generating local color feature vectors;
s42: for an abnormal exposure area, firstly, calculating detail information quantity by using an information entropy method, and then measuring local gradient change by using the local contrast of the d-th pixel obtained by the following formula;
wherein idIs the gray value of the d-th pixel point,
the gray value of the p-th neighbor pixel point in the peripheral local area of the d-th pixel point is shown, and p is the number of the local neighbor pixel points.
5. The method for evaluating the quality of a tone-mapped image partitioned based on exposure properties according to claim 1, wherein said S13 further comprises: performing local texture measurement and calculation by using a centrosymmetric local binary pattern operator, and generating 16 local structure feature vectors by each tone mapping image; wherein the centrosymmetric local binary pattern operator CS-LBP is:
μ(a，b)＝a*21+b*20；
where N is the total number of neighbors of the pixel involved, mcGray value of the central pixel, mi(i ═ 0, 1., (N/2) -1) is the gray scale value of the neighboring pixel.
6. The method for evaluating the quality of a tone-mapped image partitioned based on exposure properties according to claim 1, wherein said S14 further comprises: converting the L channel of the tone mapping image into a gray level co-occurrence matrix, and extracting three characteristics which embody the thin edge characteristics: energy E, correlation C, and uniformity V; wherein,
μ and σ denote the mean and variance of the pixel, where,
where G (x, y) is the x-th row of the gray level co-occurrence matrix, the element of the second column, and k is the scale range of the gray level domain.
7. The method for evaluating the quality of a tone-mapped image partitioned based on exposure properties according to claim 1, wherein said S15 further comprises:
s71: extracting the edge of each tone mapping image by using a Canny operator;
s72: calculating the average value H of the brightness of small blocks with preset sizes around the edge pixel points extracted in the step S711Variance H2Degree of sum deviation H3To measure the brightness of the image edge and the degree of divergence of the edge region; wherein,
H1＝E(he)；
H2＝E(he 2)-E(he)2；
8. a system for evaluating the quality of a tone-mapped image partitioned based on exposure attributes, comprising: the device comprises a partitioning unit, a partitioning local feature extraction unit, a local structure feature extraction unit, a surface structure feature extraction unit, an edge structure feature extraction unit, a model training unit and a test unit; wherein,
the partitioning unit is used for partitioning the tone mapping image into a normal exposure area and an abnormal exposure area according to an exposure value;
the regional local feature extraction unit is used for dividing the tone mapping image into a normal exposure region and an abnormal exposure region according to an exposure value;
the local structure feature extraction unit is used for extracting local texture features of the tone mapping image;
the surface structure feature extraction unit is used for extracting texture features of the tone mapping image;
the edge structure characteristic extraction unit is used for measuring the brightness of the tone mapping image edge and the divergence degree of the edge area;
the model training unit is used for combining the features extracted by the partition local feature extraction unit, the local structure feature extraction unit, the surface structure feature extraction unit and the edge structure feature extraction unit into an integral feature vector, and training a support vector regressor by continuous iteration by taking corresponding subjective quality scores provided in an image database as labels;
and the test unit is used for inputting the characteristic value of the test data in the data set into the support vector regression trained by the model training unit to obtain the predicted objective quality score of the image.
9. A quality evaluation terminal for tone-mapped images, comprising: memory, processor and computer program stored on the memory and executable on the processor, characterized in that the processor, when executing the program, is adapted to perform the method of any of claims 1-7.
10. A computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, is adapted to carry out the method of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110494547.1A CN113362354B (en) | 2021-05-07 | 2021-05-07 | Quality evaluation method, system, terminal and storage medium for tone mapping image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110494547.1A CN113362354B (en) | 2021-05-07 | 2021-05-07 | Quality evaluation method, system, terminal and storage medium for tone mapping image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113362354A true CN113362354A (en) | 2021-09-07 |
| CN113362354B CN113362354B (en) | 2024-04-30 |
Family
ID=77525861
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110494547.1A Active CN113362354B (en) | 2021-05-07 | 2021-05-07 | Quality evaluation method, system, terminal and storage medium for tone mapping image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113362354B (en) |
Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| GB201712333D0 (en) * | 2017-08-01 | 2017-09-13 | Jaguar Land Rover Ltd | Image processor and method for image processing |
| CN107172418A (en) * | 2017-06-08 | 2017-09-15 | 宁波大学 | A kind of tone scale map image quality evaluating method analyzed based on exposure status |
| CN107316279A (en) * | 2017-05-23 | 2017-11-03 | 天津大学 | Low light image Enhancement Method with regularization model is mapped based on tone |
| CN109214420A (en) * | 2018-07-27 | 2019-01-15 | 北京工商大学 | The high texture image classification method and system of view-based access control model conspicuousness detection |
| CN109377472A (en) * | 2018-09-12 | 2019-02-22 | 宁波大学 | A kind of eye fundus image quality evaluating method |
| CN110046673A (en) * | 2019-04-25 | 2019-07-23 | 上海大学 | No reference tone mapping graph image quality evaluation method based on multi-feature fusion |
| US20200007732A1 (en) * | 2018-07-02 | 2020-01-02 | Altek Corporation | Image processing method and electronic device |
| CN110717892A (en) * | 2019-09-18 | 2020-01-21 | 宁波大学 | Tone mapping image quality evaluation method |
| CN110910347A (en) * | 2019-10-18 | 2020-03-24 | 宁波大学 | Image segmentation-based tone mapping image no-reference quality evaluation method |
| CN111489333A (en) * | 2020-03-31 | 2020-08-04 | 天津大学 | No-reference night natural image quality evaluation method |
| CN112132774A (en) * | 2019-07-29 | 2020-12-25 | 方玉明 | Quality evaluation method of tone mapping image |
| CN112330648A (en) * | 2020-11-12 | 2021-02-05 | 深圳大学 | No-reference image quality evaluation method and device based on semi-supervised learning |
| CN112734650A (en) * | 2019-10-14 | 2021-04-30 | 武汉科技大学 | Virtual multi-exposure fusion based uneven illumination image enhancement method |
-
2021
- 2021-05-07 CN CN202110494547.1A patent/CN113362354B/en active Active
Patent Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107316279A (en) * | 2017-05-23 | 2017-11-03 | 天津大学 | Low light image Enhancement Method with regularization model is mapped based on tone |
| CN107172418A (en) * | 2017-06-08 | 2017-09-15 | 宁波大学 | A kind of tone scale map image quality evaluating method analyzed based on exposure status |
| GB201712333D0 (en) * | 2017-08-01 | 2017-09-13 | Jaguar Land Rover Ltd | Image processor and method for image processing |
| US20200007732A1 (en) * | 2018-07-02 | 2020-01-02 | Altek Corporation | Image processing method and electronic device |
| CN109214420A (en) * | 2018-07-27 | 2019-01-15 | 北京工商大学 | The high texture image classification method and system of view-based access control model conspicuousness detection |
| CN109377472A (en) * | 2018-09-12 | 2019-02-22 | 宁波大学 | A kind of eye fundus image quality evaluating method |
| CN110046673A (en) * | 2019-04-25 | 2019-07-23 | 上海大学 | No reference tone mapping graph image quality evaluation method based on multi-feature fusion |
| CN112132774A (en) * | 2019-07-29 | 2020-12-25 | 方玉明 | Quality evaluation method of tone mapping image |
| CN110717892A (en) * | 2019-09-18 | 2020-01-21 | 宁波大学 | Tone mapping image quality evaluation method |
| CN112734650A (en) * | 2019-10-14 | 2021-04-30 | 武汉科技大学 | Virtual multi-exposure fusion based uneven illumination image enhancement method |
| CN110910347A (en) * | 2019-10-18 | 2020-03-24 | 宁波大学 | Image segmentation-based tone mapping image no-reference quality evaluation method |
| CN111489333A (en) * | 2020-03-31 | 2020-08-04 | 天津大学 | No-reference night natural image quality evaluation method |
| CN112330648A (en) * | 2020-11-12 | 2021-02-05 | 深圳大学 | No-reference image quality evaluation method and device based on semi-supervised learning |
Non-Patent Citations (2)
| Title |
|---|
| ER-YIN SU: ""Combining Global and Local Feature Analyses for Quality Evaluation of Tone-Mapped HDR Images"", 《IEEE ACCESS》, 31 December 2018 (2018-12-31) * |
| 李靖: ""基于局部二值模式的人脸图像特征提取研究"", 《中国博士学位论文全文数据库 (信息科技辑)》, 15 January 2019 (2019-01-15), pages 138 - 87 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113362354B (en) | 2024-04-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP6792842B2 (en) | Visual inspection equipment, conversion data generation equipment, and programs | |
| CN110046673B (en) | No-reference tone mapping image quality evaluation method based on multi-feature fusion | |
| Fu et al. | A multi-task network for joint specular highlight detection and removal | |
| CN108428227B (en) | No-reference image quality evaluation method based on full convolution neural network | |
| Gu et al. | No-reference quality metric of contrast-distorted images based on information maximization | |
| Kundu et al. | No-reference quality assessment of tone-mapped HDR pictures | |
| Bahrami et al. | A fast approach for no-reference image sharpness assessment based on maximum local variation | |
| Celik | Spatial mutual information and PageRank-based contrast enhancement and quality-aware relative contrast measure | |
| US20120155765A1 (en) | Image quality assessment | |
| CN108961180B (en) | Infrared image enhancement method and system | |
| CN107146220B (en) | A kind of universal non-reference picture quality appraisement method | |
| CN107172418A (en) | A kind of tone scale map image quality evaluating method analyzed based on exposure status | |
| CN110910347B (en) | Tone mapping image non-reference quality evaluation method based on image segmentation | |
| CN105894522B (en) | A kind of more distortion objective evaluation method for quality of stereo images | |
| CN104658002A (en) | Non-reference image objective quality evaluation method | |
| Chen et al. | Perceptual quality assessment of cartoon images | |
| Fang et al. | Blind quality assessment for tone-mapped images by analysis of gradient and chromatic statistics | |
| Steffens et al. | Cnn based image restoration: Adjusting ill-exposed srgb images in post-processing | |
| Kundu et al. | No-reference image quality assessment for high dynamic range images | |
| CN111641822B (en) | Method for evaluating quality of repositioning stereo image | |
| CN105243385A (en) | Unsupervised learning based image quality evaluation method | |
| CN106210711B (en) | One kind is without with reference to stereo image quality evaluation method | |
| Dwivedi et al. | Single image dehazing using extended local dark channel prior | |
| Jiang et al. | Blind quality evaluator of tone-mapped HDR and multi-exposure fused images for electronic display | |
| CN112132774A (en) | Quality evaluation method of tone mapping image |
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 |