CN116958122B - SAR image evaluation method, SAR image evaluation device, SAR image evaluation equipment and readable storage medium - Google Patents
SAR image evaluation method, SAR image evaluation device, SAR image evaluation equipment and readable storage medium Download PDFInfo
- Publication number
- CN116958122B CN116958122B CN202311076893.3A CN202311076893A CN116958122B CN 116958122 B CN116958122 B CN 116958122B CN 202311076893 A CN202311076893 A CN 202311076893A CN 116958122 B CN116958122 B CN 116958122B
- Authority
- CN
- China
- Prior art keywords
- sar image
- evaluation
- training sample
- preset
- quality
- 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.)
- Active
Links
- 238000011156 evaluation Methods 0.000 title claims abstract description 97
- 238000012549 training Methods 0.000 claims abstract description 149
- 238000013210 evaluation model Methods 0.000 claims abstract description 49
- 238000000034 method Methods 0.000 claims abstract description 22
- 238000013441 quality evaluation Methods 0.000 claims description 19
- 230000009466 transformation Effects 0.000 claims description 17
- 238000012545 processing Methods 0.000 claims description 12
- 238000004364 calculation method Methods 0.000 claims description 7
- 238000012360 testing method Methods 0.000 claims description 6
- 238000010606 normalization Methods 0.000 claims description 5
- 238000004458 analytical method Methods 0.000 claims description 3
- 230000000694 effects Effects 0.000 abstract description 2
- 238000004891 communication Methods 0.000 description 9
- 238000010586 diagram Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000003384 imaging method Methods 0.000 description 2
- 230000002093 peripheral effect Effects 0.000 description 2
- GNFTZDOKVXKIBK-UHFFFAOYSA-N 3-(2-methoxyethoxy)benzohydrazide Chemical compound COCCOC1=CC=CC(C(=O)NN)=C1 GNFTZDOKVXKIBK-UHFFFAOYSA-N 0.000 description 1
- FGUUSXIOTUKUDN-IBGZPJMESA-N C1(=CC=CC=C1)N1C2=C(NC([C@H](C1)NC=1OC(=NN=1)C1=CC=CC=C1)=O)C=CC=C2 Chemical compound C1(=CC=CC=C1)N1C2=C(NC([C@H](C1)NC=1OC(=NN=1)C1=CC=CC=C1)=O)C=CC=C2 FGUUSXIOTUKUDN-IBGZPJMESA-N 0.000 description 1
- YTAHJIFKAKIKAV-XNMGPUDCSA-N [(1R)-3-morpholin-4-yl-1-phenylpropyl] N-[(3S)-2-oxo-5-phenyl-1,3-dihydro-1,4-benzodiazepin-3-yl]carbamate Chemical compound O=C1[C@H](N=C(C2=C(N1)C=CC=C2)C1=CC=CC=C1)NC(O[C@H](CCN1CCOCC1)C1=CC=CC=C1)=O YTAHJIFKAKIKAV-XNMGPUDCSA-N 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000000802 evaporation-induced self-assembly Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000001303 quality assessment method Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
- G06V10/774—Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting
-
- 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
- G06T2207/10032—Satellite or aerial image; Remote sensing
- G06T2207/10044—Radar 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- Computing Systems (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Databases & Information Systems (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Multimedia (AREA)
- Quality & Reliability (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
The application discloses a SAR image evaluation method, a device, equipment and a readable storage medium, wherein the method comprises the following steps: acquiring an SAR image to be evaluated; dividing the SAR image into a preset number of target areas, and respectively extracting characteristic information in each target area; inputting the characteristic information into a preset evaluation model for the evaluation model to evaluate the quality of the characteristic information, so as to obtain an evaluation result; the evaluation model is obtained by performing iterative training on a model to be trained based on a feature training sample and a quality label of the feature training sample; and determining SAR images with quality meeting preset standards according to the evaluation results. According to the SAR image evaluation method and device, the SAR image is obtained, the characteristic information is evaluated by using the preset evaluation model according to the characteristic information of the target area of the SAR image, and therefore the effect of accurately evaluating the SAR image is achieved.
Description
Technical Field
The present application relates to the field of synthetic aperture radar technologies, and in particular, to a method, an apparatus, a device, and a readable storage medium for evaluating SAR images.
Background
Synthetic aperture radar (SAR, SYNTHETIC APERTURE RADAR) is widely used in the field of target detection and identification technologies. The unmanned aerial vehicle SAR system with real-time imaging capability can be used for realizing low-delay real-time imaging on the target, effectively and dynamically monitoring the target, and has high practical application value in the fields of emergency, disasters and the like.
After the corresponding SAR image is formed through SAR echo simulation, the actual condition of the corresponding area of the current SAR image is determined by interpretation and recognition of the image by the related personnel, and the quality of the SAR image directly influences the recognition and judgment of the related personnel on the SAR image, so that the quality of the SAR image needs to be determined before the related personnel recognize the SAR image.
When the quality of the SAR image is evaluated, the quality of the SAR image can be evaluated in the same mode due to negligence, the SAR images corresponding to different characteristic types cannot be evaluated respectively, the evaluation accuracy is low, and the accuracy of the follow-up related personnel in recognizing the SAR image is low.
Disclosure of Invention
In view of the above, the present application provides a SAR image evaluation method, apparatus, device, and readable storage medium, which aim to improve accuracy of SAR image quality evaluation.
To achieve the above object, the present application provides a SAR image evaluation method comprising the steps of:
Acquiring an SAR image to be evaluated;
dividing the SAR image into a preset number of target areas, and respectively extracting characteristic information in each target area;
Inputting the characteristic information into a preset evaluation model for the evaluation model to evaluate the quality of the characteristic information, so as to obtain an evaluation result; the evaluation model is obtained by performing iterative training on a model to be trained based on a feature training sample and a quality label of the feature training sample;
And determining SAR images with quality meeting preset standards according to the evaluation results.
Exemplary, after the step of dividing the SAR image into a preset number of target areas and extracting the feature information in each target area, the method includes:
acquiring a feature training sample and a quality label of the feature training sample;
Determining a quality evaluation index of the feature training sample with the quality label based on the feature training sample and the quality label, and determining an index weight of the quality evaluation index;
And performing iterative training on a preset model to be trained based on the characteristic training sample and the quality label to obtain an evaluation model meeting the accuracy condition.
The step of iteratively training the preset model to be trained based on the feature training sample and the quality label to obtain a test prediction model meeting the accuracy condition includes:
Inputting the feature training sample and the index weight into the model to be trained to obtain a predicted feature evaluation value;
Performing difference calculation on the predicted characteristic evaluation value and the quality label of the characteristic training sample to obtain an error result;
Based on the error result, judging whether the error result meets an error standard indicated by a preset error threshold range;
And if the error result does not meet the error standard indicated by the preset error threshold range, returning to the step of inputting the feature training sample and the index weight into the model to be trained to obtain a predicted feature evaluation value, and stopping training until the training error result meets the error standard indicated by the preset error threshold range to obtain the evaluation model.
Illustratively, before the step of obtaining the feature training samples and the quality labels of the feature training samples, the method includes:
determining characteristic information of the training sample, and generating a histogram of the training sample corresponding to the complexity according to the characteristic information;
according to the histogram, counting the complex value of each target area corresponding to the characteristic information;
calculating the average value of the sum of the complex values, and taking the average value as a standard value;
and determining the region type corresponding to each training sample according to the standard value, and generating quality labels corresponding to different training samples according to the region type and the characteristic information.
The step of determining the feature information of the training sample and generating a histogram of the training sample corresponding to the complexity according to the feature information includes:
resolving the resolution, the integrity and the fuzzy ratio in the characteristic information, and carrying out normalization processing on the resolution, the integrity and the fuzzy ratio to obtain a processed initial value;
And calculating the sum of the initial values according to preset complex weights, and generating a histogram of the training sample corresponding to the complexity according to the sum of the initial values.
Illustratively, the step of dividing the SAR image into a preset number of target areas includes:
Acquiring a preset number of target points to be evaluated;
The target point positions are used as coordinate centers, and the preset number of initial areas are generated;
and if the overlapping area exists between the initial areas, converting the plurality of initial areas with the overlapping areas into a target area, and taking the initial areas without the overlapping areas as the target area.
The step of determining, according to the evaluation result, the SAR image with quality meeting the preset standard includes:
If the SAR image does not accord with the preset standard, analyzing the target area through Curvelet transformation to obtain a high-frequency sub-band and a low-frequency sub-band;
According to a preset gain coefficient, the high-frequency sub-band and the low-frequency sub-band are adjusted, curvelet inverse transformation is carried out on the adjusted sub-bands, a denoised target area is obtained, and reevaluation is carried out on the denoised target area;
And if the SAR image after the reevaluation still does not meet the preset standard, determining that the SAR image is an image with unqualified quality.
For achieving the above object, the present application also provides an SAR image evaluation device, including:
The acquisition module is used for acquiring the SAR image to be evaluated;
The processing module is used for dividing the SAR image into a preset number of target areas and extracting characteristic information in each target area respectively;
the input module is used for inputting the characteristic information into a preset evaluation model so as to enable the evaluation model to evaluate the quality of the characteristic information and obtain an evaluation result; the evaluation model is obtained by performing iterative training on a model to be trained based on a feature training sample and a quality label of the feature training sample;
And the determining module is used for determining SAR images with the quality meeting the preset standard according to the evaluation result.
For achieving the above object, the present application also provides an SAR image evaluation device including: the SAR image system comprises a memory, a processor and an SAR image assessment program stored on the memory and capable of running on the processor, wherein the SAR image assessment program is configured to realize the steps of the SAR image assessment method.
Illustratively, to achieve the above object, the present application also provides a computer-readable storage medium having stored thereon a SAR image evaluation program which, when executed by a processor, implements the steps of the SAR image evaluation method as described above.
Compared with the condition that the accuracy of an obtained evaluation result is low in the related art, the SAR image quality is evaluated based on the related evaluation method, and the SAR image to be evaluated is acquired; dividing the SAR image into a preset number of target areas, and respectively extracting characteristic information in each target area; inputting the characteristic information into a preset evaluation model for the evaluation model to evaluate the quality of the characteristic information, so as to obtain an evaluation result; the evaluation model is obtained by performing iterative training on a model to be trained based on a feature training sample and a quality label of the feature training sample; according to the evaluation result, determining SAR images with quality meeting the preset standard, namely dividing the SAR images into a preset number of target areas, respectively extracting the characteristic information of the target areas, inputting the characteristic information into an evaluation model, evaluating the quality of the characteristic information through the evaluation model, realizing that the evaluation model evaluates the SAR images with quality meeting the preset standard according to different characteristic information in different modes, and determining the SAR images with quality meeting the preset standard according to the evaluation result obtained by the evaluation, so that the evaluation model can accurately correspond to the image characteristics of each target area in the SAR images to evaluate, thereby ensuring the accuracy of evaluating the SAR images, further improving the quality of SAR images interpreted by related personnel, and improving the accuracy and efficiency of image interpretation by related personnel.
Drawings
FIG. 1 is a flowchart of a first embodiment of the SAR image evaluation method of the present application;
FIG. 2 is a flowchart of another embodiment of the SAR image evaluation method according to the present disclosure;
fig. 3 is a schematic structural diagram of a hardware running environment according to an embodiment of the present application.
The achievement of the objects, functional features and advantages of the present application will be further described with reference to the accompanying drawings, in conjunction with the embodiments.
Detailed Description
It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the application.
The application provides a SAR image evaluation method, referring to FIG. 1, FIG. 1 is a flow chart of a first embodiment of the SAR image evaluation method of the present application.
The embodiments of the present application provide embodiments of SAR image evaluation methods, it should be noted that although a logical order is shown in the flow chart, in some cases the steps shown or described may be performed in a different order than here. For convenience of description, each step of performing a subject description SAR image evaluation method is omitted below, the SAR image evaluation method including:
Step S110: acquiring an SAR image to be evaluated;
Before related personnel read and identify the SAR image, the quality of the SAR image is required to be assessed first, and partial SAR image with poor quality is screened out, so that the subsequent related personnel can read the image with high quality, the efficiency of reading is improved, and the reading time is prevented from being used for reading the image with poor quality.
The SAR image which is not subjected to quality evaluation at present is the SAR image to be evaluated.
Step S120: dividing the SAR image into a preset number of target areas, and respectively extracting characteristic information in each target area;
The preset number can be determined according to the content of SAR images read by related personnel in actual need, the content is complex and the distribution is scattered, a plurality of target areas can be set, the preset number is the number of limiting target areas, the number is determined according to the content of SAR images, and the number is at least two.
Illustratively, the step of dividing the SAR image into a preset number of target areas includes:
Step a: acquiring a preset number of target points to be evaluated;
the target point to be evaluated can be set according to related personnel and input into the system, and the selection of the target point is mainly related to the content of the area mainly displayed by the SAR image.
Step b: the target point positions are used as coordinate centers, and the preset number of initial areas are generated;
Step c: and if the overlapping area exists between the initial areas, converting the plurality of initial areas with the overlapping areas into a target area, and taking the initial areas without the overlapping areas as the target area.
The target area is the area of the content to be read by the related personnel, the area can be defined in advance by the related personnel or is provided with positioning points in advance, and the SAR image is divided into a preset number of target areas according to the defined range or positioning points.
For example, when the target area is defined, the initial area may be generated directly according to the pre-defined range, if the initial area is not overlapped, the initial area is taken as the target area, or the target area may be defined according to a preset positioning point, for example, the positioning point is taken as the center, a square frame with fixed side length is generated, so as to select the target area, or a plurality of positioning points are directly included, and the corresponding target area is generated.
When there is an overlapping area between the initial areas, that is, when there is a phenomenon that the generated plurality of initial areas overlap, the overlapping areas need to be fused, for example, a target area with a larger area covered by positioning points corresponding to the plurality of initial areas is directly generated.
After demarcating and waiting to a preset number of target areas, feature information in each target area needs to be extracted, for example: the corresponding information of SAR image features can be represented by light and shade contrast information, structure information, pixel point position information, resolution information, image width information, edge information and the like.
Step S130: inputting the characteristic information into a preset evaluation model for the evaluation model to evaluate the quality of the characteristic information, so as to obtain an evaluation result; the evaluation model is obtained by performing iterative training on a model to be trained based on a feature training sample and a quality label of the feature training sample;
Inputting the characteristic information into a preset evaluation model, and performing quality evaluation on the characteristic information through the preset evaluation model so as to obtain an evaluation result of corresponding quality evaluation.
The preset evaluation model is obtained after iterative training is performed on the model to be trained based on the feature training sample and the quality label of the feature training sample, namely, the preset evaluation model can classify feature information and divide a corresponding target area into a complex area and a simple area according to the complexity of the feature information, wherein the feature information of the simple area is low, parameters such as the integrity, the definition and the resolution of the feature information are high, the quality of the simple area can be evaluated without additional processing, the feature information of the complex area is high, interference is high, and multiple contents such as scattering type, light-dark contrast and the like are needed to be combined when the quality of the complex area is evaluated, so that the feature information corresponding to the complex area is evaluated.
Exemplary, after the step of dividing the SAR image into a preset number of target areas and extracting the feature information in each target area, the method includes:
Step d: acquiring a feature training sample and a quality label of the feature training sample;
The feature training sample is, for example, a feature sample for training, including a certain amount of feature information.
Illustratively, the quality label of the feature training sample is an evaluation result corresponding to the quality evaluation of the feature training sample.
Step e: determining a quality evaluation index of the feature training sample with the quality label based on the feature training sample and the quality label, and determining an index weight of the quality evaluation index;
For example, different evaluation indexes can be set for different types of target areas, for example, only the resolution of a simple area is evaluated, the resolution, the blur ratio, the integrity and other indexes of a complex area need to be evaluated, the index weight corresponding to the quality evaluation index is the weight corresponding to the above index, and different indexes are set according to the different target areas.
The weight of the index other than the resolution may be set to 0 for the simple region, that is, only the resolution may be evaluated when the simple region is evaluated.
Wherein, for the complex region, the corresponding weights can be set corresponding to the three indexes respectively.
Step f: and performing iterative training on a preset model to be trained based on the characteristic training sample and the quality label to obtain an evaluation model meeting the accuracy condition.
The method comprises the steps of carrying out iterative training on a preset model to be trained based on the characteristic training sample, the quality label and the characteristic weight to obtain a test prediction model meeting accuracy conditions, and carrying out quality assessment on a target area.
The step of iteratively training the preset model to be trained based on the feature training sample and the quality label to obtain a test prediction model meeting the accuracy condition includes:
step g: inputting the feature training sample and the index weight into the model to be trained to obtain a predicted feature evaluation value;
The feature training sample and the index weight are input to the model to be trained to obtain a predicted feature evaluation value, wherein the predicted feature evaluation value is obtained by performing a prediction test on the model in training.
Step h: performing difference calculation on the predicted characteristic evaluation value and the quality label of the characteristic training sample to obtain an error result;
The difference calculation is performed on the predicted feature evaluation value and the quality label of the feature training sample, that is, whether the result obtained by the model in training is consistent with the known result is verified, and the difference calculation between the results is performed to obtain an error result.
Step i: based on the error result, judging whether the error result meets an error standard indicated by a preset error threshold range;
for example, since the result after model training has an error with the actual result, the error result is allowed to be within the preset error threshold range, so as to further judge whether the error result meets the error standard indicated by the preset error threshold range.
Step j: and if the error result does not meet the error standard indicated by the preset error threshold range, returning to the step of inputting the feature training sample and the index weight into the model to be trained to obtain a predicted feature evaluation value, and stopping training until the training error result meets the error standard indicated by the preset error threshold range to obtain the evaluation model.
If the error result does not meet the error standard indicated by the preset error threshold range, the model is shown to have overlarge error in the training, and the feature training sample and the index weight are returned to be input to the model to be trained to obtain the predicted feature evaluation value, namely iterative training is performed until the training error result meets the error standard indicated by the preset error threshold range, training is stopped, the test prediction model meeting the precision condition is obtained, and accuracy of model evaluation is improved.
Compared with the condition that the accuracy of an obtained evaluation result is low in the related art, the SAR image quality is evaluated based on the related evaluation method, and the SAR image to be evaluated is acquired; dividing the SAR image into a preset number of target areas, and respectively extracting characteristic information in each target area; inputting the characteristic information into a preset evaluation model for the evaluation model to evaluate the quality of the characteristic information, so as to obtain an evaluation result; the evaluation model is obtained by performing iterative training on a model to be trained based on a feature training sample and a quality label of the feature training sample; according to the evaluation result, determining SAR images with quality meeting the preset standard, namely dividing the SAR images into a preset number of target areas, respectively extracting the characteristic information of the target areas, inputting the characteristic information into an evaluation model, evaluating the quality of the characteristic information through the evaluation model, realizing that the evaluation model evaluates the SAR images with quality meeting the preset standard according to different characteristic information in different modes, and determining the SAR images with quality meeting the preset standard according to the evaluation result obtained by the evaluation, so that the evaluation model can accurately correspond to the image characteristics of each target area in the SAR images to evaluate, thereby ensuring the accuracy of evaluating the SAR images, further improving the quality of SAR images interpreted by related personnel, and improving the accuracy and efficiency of image interpretation by related personnel.
An embodiment is presented based on the first embodiment of the SAR image evaluation method of the present application, in which, referring to fig. 2, the method comprises:
Step S210: determining characteristic information of the training sample, and generating a histogram of the training sample corresponding to the complexity according to the characteristic information;
When generating the histogram of the complexity degree, the complexity degree of the feature information corresponding to the target region (each training sample may be regarded as the extracted target region, and the target region is used for replacing the training sample to describe below) is comprehensively considered, and the complexity degree includes the definition of the feature information (the integrity of the feature information, the resolution of the feature information, the blur ratio, the interference degree suffered by the feature information, and the like), the complexity of the structural feature in the feature information, and the like.
The step of determining the feature information of the training sample and generating a histogram of the training sample corresponding to the complexity according to the feature information includes:
Step k: resolving the resolution, the integrity and the fuzzy ratio in the characteristic information, and carrying out normalization processing on the resolution, the integrity and the fuzzy ratio to obtain a processed initial value;
step l: and calculating the sum of the initial values according to preset complex weights, and generating a histogram of the training sample corresponding to the complexity according to the sum of the initial values.
The characteristic information at least comprises information such as resolution, integrity, fuzzy ratio and the like, and is used as index information for mainly calculating the complexity of the SAR image, and the three information is normalized, for example, the respective parameters of the resolution, the integrity and the fuzzy ratio are converted into any value within 0-1.
After the data is normalized, an initial value is obtained, a preset complex weight is used, the initial value is given weight, the sum of the initial values is calculated, and the sum of the initial values is used as the complex value of the target area, so that a histogram of the complexity corresponding to the target area is correspondingly generated.
Step S220: according to the histogram, counting the complex value of each target area corresponding to the characteristic information;
In the histogram, a complex value is generated corresponding to each target region, and the greater the complex value is, the greater the complexity corresponding to the feature information is proved, and the complex value corresponds to the complexity of the feature information.
Step S230: calculating the average value of the sum of the complex values, and taking the average value as a standard value;
after generating the histograms, respectively counting complex values in each histogram, wherein the complex values have a one-to-one correspondence with the target areas corresponding to the characteristic information. And summing the complex values, calculating a summed average value, and taking the average value as a standard value, wherein the standard value is used for distinguishing the region type of each current target region.
After calculating the average value, the region types of each target region are screened in advance according to the average value, the proportion of different region types is calculated according to the pre-screening result, if the proportion exceeds the preset proportion, the average value needs to be calculated again, and the lowest and highest complex values can be removed correspondingly during the recalculation.
Step S240: and determining the region type corresponding to each training sample according to the standard value, and generating quality labels corresponding to different training samples according to the region type and the characteristic information.
The region types of each target region can be simple regions and complex regions, when the attribute information of the region types of different training samples is generated according to the region types, the region types with preset proportion are required to be set, and the attribute of the region types is added into the quality label, namely, the final output result of the control model is different quality evaluation results aiming at the regions with different types, for example, for the simple regions, only the evaluation results corresponding to part of indexes are output, and for the complex regions, the evaluation results corresponding to more indexes than the simple regions are required to be output.
The preset ratio may be set according to practical situations, specifically, the ratio of the number of simple areas to the number of complex areas is defined as 1:2, avoiding the problem that the subsequent evaluation efficiency is low due to the excessive number of the target areas of the classified and complex areas, namely controlling the number of the complex areas obtained by model classification. And determining the region type corresponding to each target region according to the standard value, for example, determining the target region as a complex region if the complex value of the characteristic information corresponding to each target region is larger than the standard value, and determining the target region as a simple region if the complex value of the characteristic information corresponding to each target region is smaller than the standard value.
In addition, the feature information is used as the feature of the SAR image, and the region type of the target region corresponding to the feature information can be determined according to the number of the feature information or the complexity of the feature information, and the region type can be divided into a simple type region and a complex type region.
In the simple region, the number of the feature information contained therein is small, the complexity of the feature information is low, and in the complex region, the number of the feature information contained therein is large, and the complexity of the feature information is high.
For example, the light and shade contrast ratio of the target area is low, so that the structural characteristics of the target area are not obvious, the target area can be used as a complex area, or the resolution ratio of the target area is low or the edge information in the target area is complex, the most accurate information can not be directly extracted, and the target area can be used as the complex area.
Different areas need to be evaluated in different ways, for example, the quality of the target area cannot be accurately evaluated when the complex area is evaluated due to the fact that the characteristic information of the complex area is not obvious or the characteristic information of the complex area is complex, so that the complex area needs to be further processed correspondingly, and the complex area can be accurately evaluated.
The step of determining, according to the evaluation result, the SAR image with quality meeting the preset standard includes:
Step m: if the SAR image does not accord with the preset standard, analyzing the target area through Curvelet transformation to obtain a high-frequency sub-band and a low-frequency sub-band;
Step n: according to a preset gain coefficient, the high-frequency sub-band and the low-frequency sub-band are adjusted, curvelet inverse transformation is carried out on the adjusted sub-bands, a denoised target area is obtained, and reevaluation is carried out on the denoised target area;
Step o: and if the SAR image after the reevaluation still does not meet the preset standard, determining that the SAR image is an image with unqualified quality.
When the SAR image does not meet the preset standard, the SAR image has a possibility that the target area part in the SAR image is a complex area, and the characteristic information of the complex area has serious interference, and at the moment, the SAR image needs to be correspondingly processed, so that the interference and noise are reduced.
Curvelet transformation is a multi-scale geometric analysis method with multi-directionality, which is developed on the basis of wavelet transformation. The method is suitable for analyzing the curve or linear edge characteristics of the two-dimensional image, and has higher approximation precision and better sparse expression capability. Based on the image denoising and enhancement thought of Curvelet transformation, the high-frequency small coefficients of the Curvelet transformation domain are suppressed by adopting a multi-threshold method or by designing a nonlinear enhancement function, and the large coefficients are enhanced. But the Curvelet transformation has better expression on information such as straight lines, curves and the like in the image, and can obtain more ideal denoising and enhancement effects than wavelet transformation.
Specifically, curvelet transformation can be used to decompose the image to obtain a low-frequency subband and high-frequency subbands with different scales and different directions; multiplying the low-frequency subband coefficient by a small gain coefficient to inhibit the background information of the image; meanwhile, a directional selective high-frequency enhancement function needs to be designed, and a larger gain multiple and a smaller threshold value are set for a high-frequency subband close to the horizontal direction; setting a smaller gain multiple and a larger threshold for a high frequency subband close to the vertical direction; and finally, performing Curvelet inverse transformation on the processed coefficients to obtain an enhanced image.
Wherein the number of decomposition scales is chosen to be 4-8 (including low frequency subbands), and for the second coarse scale level the number of directions is chosen to be 8 or 16.
Wherein the preset gain coefficient takes a value of 0.6-1.
In this embodiment, feature information of the training sample is determined, and a histogram of the training sample corresponding to the complexity degree is generated according to the feature information; according to the histogram, counting the complex value of each training sample corresponding to the characteristic information; calculating the average value of the sum of the complex values, and taking the average value as a standard value; and determining the region type corresponding to each training sample according to the standard value, and generating quality labels corresponding to different training samples according to the region type and the characteristic information, namely, according to different scenes for carrying out quality evaluation on the image, and simultaneously according to the attribute of the corresponding region type in the quality label, so as to influence the accuracy of the model in the final quality evaluation of the image.
In addition, the present application also provides a SAR image evaluation device, which includes:
The acquisition module is used for acquiring the SAR image to be evaluated;
The processing module is used for dividing the SAR image into a preset number of target areas and extracting characteristic information in each target area respectively;
the input module is used for inputting the characteristic information into a preset evaluation model so as to enable the evaluation model to evaluate the quality of the characteristic information and obtain an evaluation result; the evaluation model is obtained by performing iterative training on a model to be trained based on a feature training sample and a quality label of the feature training sample;
And the determining module is used for determining SAR images with the quality meeting the preset standard according to the evaluation result.
Illustratively, the apparatus further comprises:
The sample acquisition module is used for acquiring the characteristic training samples and quality labels of the characteristic training samples;
A sample determining module, configured to determine a quality evaluation index of a feature training sample with the quality label based on the feature training sample and the quality label, and determine an index weight of the quality evaluation index;
And the sample training module is used for carrying out iterative training on a preset model to be trained based on the characteristic training sample and the quality label to obtain an evaluation model meeting the accuracy condition.
Illustratively, the sample training module includes:
The input sub-module is used for inputting the feature training sample and the index weight into the model to be trained to obtain a predicted feature evaluation value;
The first computing sub-module is used for carrying out difference computation on the predicted characteristic evaluation value and the quality label of the characteristic training sample to obtain an error result;
The judging submodule is used for judging whether the error result meets an error standard indicated by a preset error threshold range or not based on the error result;
The iteration sub-module is used for returning to the step of inputting the characteristic training sample and the index weight into the model to be trained to obtain a predicted characteristic evaluation value if the error result does not meet the error standard indicated by the preset error threshold range, and stopping training until the training error result meets the error standard indicated by the preset error threshold range to obtain the evaluation model;
The first generation sub-module is used for determining the characteristic information of the training sample and generating a histogram of the corresponding complexity degree of the training sample according to the characteristic information;
The statistics sub-module is used for counting the complex value of each target area corresponding to the characteristic information according to the histogram;
the second calculation sub-module is used for calculating the average value of the sum of the complex values and taking the average value as a standard value;
And the determining submodule is used for determining the region type corresponding to each training sample according to the standard value and generating quality labels corresponding to different training samples according to the region type and the characteristic information.
Illustratively, the first generating sub-module includes:
the analysis unit is used for analyzing the resolution, the integrity and the fuzzy ratio in the characteristic information, and carrying out normalization processing on the resolution, the integrity and the fuzzy ratio to obtain a processed initial value;
And the generating unit is used for calculating the sum of the initial values according to preset complex weights and generating a histogram of the training sample corresponding to the complexity degree according to the sum of the initial values.
Illustratively, the processing module includes:
The acquisition sub-module is used for acquiring a preset number of target point positions to be evaluated;
The second generation submodule is used for taking the target point positions as coordinate centers and generating the preset number of initial areas;
And the conversion sub-module is used for converting a plurality of initial areas with the overlapping areas into a target area if the overlapping areas exist between the initial areas, and taking the initial areas without the overlapping areas as the target areas.
Illustratively, the apparatus further comprises:
the transformation module is used for analyzing the target area through Curvelet transformation if the SAR image does not accord with the preset standard, so as to obtain a high-frequency sub-band and a low-frequency sub-band;
The adjusting module is used for adjusting the high-frequency sub-band and the low-frequency sub-band according to a preset gain coefficient, performing Curvelet inverse transformation on the adjusted sub-bands to obtain a denoised target area, and reevaluating the denoised target area;
And the classification module is used for determining the SAR image to be an image with unqualified quality if the SAR image after the heavy evaluation still does not meet the preset standard.
The specific implementation manner of the SAR image evaluation device is basically the same as that of each embodiment of the SAR image evaluation method, and is not repeated here.
In addition, the application also provides SAR image evaluation equipment. As shown in fig. 3, fig. 3 is a schematic structural diagram of a hardware running environment according to an embodiment of the present application.
By way of example, fig. 3 may be a schematic structural diagram of a hardware operating environment of the SAR image evaluation device.
As shown in fig. 3, the SAR image evaluation device may include a processor 301, a communication interface 302, a memory 303, and a communication bus 304, wherein the processor 301, the communication interface 302, and the memory 303 complete communication with each other through the communication bus 304, and the memory 303 is used for storing a computer program; the processor 301 is configured to implement the steps of the SAR image evaluation method when executing the program stored in the memory 303.
The communication bus 304 mentioned by the above SAR image evaluation device may be a peripheral component interconnect standard (PERIPHERAL COMPONENT INTERCONNECT, PCI) bus, an extended industry standard architecture (Extended Industry Standard Architecture, EISA) bus, or the like. The communication bus 304 may be divided into an address bus, a data bus, a control bus, and the like. For ease of illustration, the figures are shown with only one bold line, but not with only one bus or one type of bus.
The communication interface 302 is used for communication between the above-described SAR image evaluation device and other devices.
The Memory 303 may include a random access Memory (Random Access Memory, RMD) or may include a Non-Volatile Memory (NM), such as at least one disk Memory. Optionally, the memory 303 may also be at least one memory device located remotely from the aforementioned processor 301.
The processor 301 may be a general-purpose processor, including a central processing unit (Central Processing Unit, CPU), a network processor (Network Processor, NP), etc.; but may also be a digital signal Processor (DIGITAL SIGNAL Processor, DSP), application SPECIFIC INTEGRATED Circuit (ASIC), field-Programmable gate array (Field-Programmable GATE ARRAY, FPGA) or other Programmable logic device, discrete gate or transistor logic device, discrete hardware components.
The specific implementation manner of the SAR image evaluation device is basically the same as that of each embodiment of the SAR image evaluation method, and is not repeated here.
In addition, the embodiment of the application also provides a computer readable storage medium, wherein the computer readable storage medium stores a SAR image evaluation program, and the SAR image evaluation program realizes the steps of the SAR image evaluation method when being executed by a processor.
The specific implementation manner of the computer readable storage medium of the present application is basically the same as the above embodiments of the SAR image evaluation method, and will not be described herein.
It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or system. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or system that comprises the element.
The foregoing embodiment numbers of the present application are merely for the purpose of description, and do not represent the advantages or disadvantages of the embodiments.
From the above description of the embodiments, it will be clear to those skilled in the art that the above-described embodiment method may be implemented by means of software plus a necessary general hardware platform, but of course may also be implemented by means of hardware, but in many cases the former is a preferred embodiment. Based on such understanding, the technical solution of the present application may be embodied essentially or in a part contributing to the prior art in the form of a software product stored in a storage medium (e.g. ROM/RAM, magnetic disk, optical disk) as described above, comprising instructions for causing a terminal device (which may be a mobile phone, a computer, a server, or a network device, etc.) to perform the method according to the embodiments of the present application.
The foregoing description is only of the preferred embodiments of the present application, and is not intended to limit the scope of the application, but rather is intended to cover any equivalents of the structures or equivalent processes disclosed herein or in the alternative, which may be employed directly or indirectly in other related arts.
Claims (7)
1. A SAR image evaluation method, characterized in that the SAR image evaluation method comprises the steps of:
Acquiring an SAR image to be evaluated;
dividing the SAR image into a preset number of target areas, and respectively extracting characteristic information in each target area;
After the step of dividing the SAR image into a preset number of target areas and extracting the characteristic information in each target area, the method comprises the following steps:
acquiring a feature training sample and a quality label of the feature training sample;
Determining a quality evaluation index of the feature training sample with the quality label based on the feature training sample and the quality label, and determining an index weight of the quality evaluation index;
performing iterative training on a preset model to be trained based on the characteristic training sample and the quality label to obtain an evaluation model meeting the accuracy condition;
before the step of obtaining the feature training sample and the quality label of the feature training sample, the method comprises the following steps:
determining characteristic information of the training sample, and generating a histogram of the training sample corresponding to the complexity according to the characteristic information;
the step of determining the feature information of the training sample and generating a histogram of the training sample corresponding to the complexity according to the feature information comprises the following steps:
resolving the resolution, the integrity and the fuzzy ratio in the characteristic information, and carrying out normalization processing on the resolution, the integrity and the fuzzy ratio to obtain a processed initial value;
Calculating the sum of the initial values according to preset complex weights, and generating a histogram of the training sample corresponding to the complexity according to the sum of the initial values;
according to the histogram, counting the complex value of each training sample corresponding to the characteristic information;
calculating the average value of the sum of the complex values, and taking the average value as a standard value;
Determining the region type corresponding to each training sample according to the standard value, and generating quality labels corresponding to different training samples according to the region type and the characteristic information, wherein the region type is a simple region and a complex region, and the proportion between the simple region and the complex region does not exceed a preset proportion;
Inputting the characteristic information into a preset evaluation model for the evaluation model to evaluate the quality of the characteristic information, so as to obtain an evaluation result; the evaluation model is obtained by performing iterative training on a model to be trained based on a feature training sample and a quality label of the feature training sample;
And determining SAR images with quality meeting preset standards according to the evaluation results.
2. The SAR image evaluation method according to claim 1, wherein the step of iteratively training a preset model to be trained based on the feature training sample and the quality label to obtain a test prediction model satisfying a precision condition comprises:
Inputting the feature training sample and the index weight into the model to be trained to obtain a predicted feature evaluation value;
Performing difference calculation on the predicted characteristic evaluation value and the quality label of the characteristic training sample to obtain an error result;
Based on the error result, judging whether the error result meets an error standard indicated by a preset error threshold range;
And if the error result does not meet the error standard indicated by the preset error threshold range, returning to the step of inputting the feature training sample and the index weight into the model to be trained to obtain a predicted feature evaluation value, and stopping training until the training error result meets the error standard indicated by the preset error threshold range to obtain the evaluation model.
3. The SAR image evaluation method of claim 1, wherein the step of dividing the SAR image into a preset number of target areas comprises:
Acquiring a preset number of target points to be evaluated;
The target point positions are used as coordinate centers, and the preset number of initial areas are generated;
and if the overlapping area exists between the initial areas, converting the plurality of initial areas with the overlapping areas into a target area, and taking the initial areas without the overlapping areas as the target area.
4. The SAR image evaluation method according to claim 1, wherein after the step of determining the SAR image having a quality according to the predetermined criterion based on the evaluation result, the method comprises:
If the SAR image does not accord with the preset standard, analyzing the target area through Curvelet transformation to obtain a high-frequency sub-band and a low-frequency sub-band;
According to a preset gain coefficient, the high-frequency sub-band and the low-frequency sub-band are adjusted, curvelet inverse transformation is carried out on the adjusted sub-bands, a denoised target area is obtained, and reevaluation is carried out on the denoised target area;
And if the SAR image after the reevaluation still does not meet the preset standard, determining that the SAR image is an image with unqualified quality.
5. A SAR image evaluation device, characterized in that the SAR image evaluation device comprises:
The acquisition module is used for acquiring the SAR image to be evaluated;
The processing module is used for dividing the SAR image into a preset number of target areas and extracting characteristic information in each target area respectively;
The sample acquisition module is used for acquiring the characteristic training samples and quality labels of the characteristic training samples;
A sample determining module, configured to determine a quality evaluation index of a feature training sample with the quality label based on the feature training sample and the quality label, and determine an index weight of the quality evaluation index;
The sample training module is used for carrying out iterative training on a preset model to be trained based on the characteristic training sample and the quality label to obtain an evaluation model meeting the accuracy condition;
The first generation sub-module is used for determining the characteristic information of the training sample and generating a histogram of the corresponding complexity degree of the training sample according to the characteristic information;
The first generation sub-module includes:
the analysis unit is used for analyzing the resolution, the integrity and the fuzzy ratio in the characteristic information, and carrying out normalization processing on the resolution, the integrity and the fuzzy ratio to obtain a processed initial value;
the generating unit is used for calculating the sum of the initial values according to preset complex weights and generating a histogram of the training sample corresponding to the complexity according to the sum of the initial values;
The statistics sub-module is used for counting the complex value of each target area corresponding to the characteristic information according to the histogram;
the second calculation sub-module is used for calculating the average value of the sum of the complex values and taking the average value as a standard value;
the determining submodule is used for determining the region type corresponding to each training sample according to the standard value and generating quality labels corresponding to different training samples according to the region type and the characteristic information, wherein the region type is a simple region and a complex region, and the proportion between the simple region and the complex region does not exceed a preset proportion;
the input module is used for inputting the characteristic information into a preset evaluation model so as to enable the evaluation model to evaluate the quality of the characteristic information and obtain an evaluation result; the evaluation model is obtained by performing iterative training on a model to be trained based on a feature training sample and a quality label of the feature training sample;
And the determining module is used for determining SAR images with the quality meeting the preset standard according to the evaluation result.
6. A SAR image evaluation device, characterized in that the device comprises: a memory, a processor and a SAR image evaluation program stored on the memory and executable on the processor, the SAR image evaluation program being configured to implement the steps of the SAR image evaluation method of any one of claims 1 to 4.
7. A computer-readable storage medium, characterized in that the computer-readable storage medium has stored thereon a SAR image evaluation program which, when executed by a processor, implements the steps of the SAR image evaluation method according to any one of claims 1 to 4.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202311076893.3A CN116958122B (en) | 2023-08-24 | 2023-08-24 | SAR image evaluation method, SAR image evaluation device, SAR image evaluation equipment and readable storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202311076893.3A CN116958122B (en) | 2023-08-24 | 2023-08-24 | SAR image evaluation method, SAR image evaluation device, SAR image evaluation equipment and readable storage medium |
Publications (2)
Publication Number | Publication Date |
---|---|
CN116958122A CN116958122A (en) | 2023-10-27 |
CN116958122B true CN116958122B (en) | 2024-06-21 |
Family
ID=88447635
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202311076893.3A Active CN116958122B (en) | 2023-08-24 | 2023-08-24 | SAR image evaluation method, SAR image evaluation device, SAR image evaluation equipment and readable storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116958122B (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN118069099B (en) * | 2024-04-24 | 2024-07-09 | 北京东远润兴科技有限公司 | FPGA-based multi-matrix parallel pipeline transposed SAR imaging method, device, equipment and storage medium |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112102309A (en) * | 2020-09-27 | 2020-12-18 | 中国建设银行股份有限公司 | Method, device and equipment for determining image quality evaluation result |
CN112950581A (en) * | 2021-02-25 | 2021-06-11 | 北京金山云网络技术有限公司 | Quality evaluation method and device and electronic equipment |
Family Cites Families (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108550145B (en) * | 2018-04-11 | 2021-01-29 | 北京环境特性研究所 | SAR image quality evaluation method and device |
GB201904857D0 (en) * | 2019-04-05 | 2019-05-22 | Univ Oxford Innovation Ltd | Quality assessment in video endosscopy |
CN110703215B (en) * | 2019-10-16 | 2022-09-06 | 电子科技大学 | Airborne SAR imaging quality evaluation method based on support vector machine |
CN112700408B (en) * | 2020-12-28 | 2023-09-08 | ***股份有限公司 | Model training method, image quality evaluation method and device |
CN116091900A (en) * | 2021-10-29 | 2023-05-09 | 华为技术有限公司 | Image processing method, device and storage medium |
CN114091969A (en) * | 2021-12-01 | 2022-02-25 | 上海明胜品智人工智能科技有限公司 | Meal quality evaluation method and device, electronic equipment and readable storage medium |
CN114913159A (en) * | 2022-05-23 | 2022-08-16 | 中国科学院深圳先进技术研究院 | Ultrasonic image quality evaluation method, model training method and electronic equipment |
CN115631399A (en) * | 2022-10-25 | 2023-01-20 | 深圳须弥云图空间科技有限公司 | Training method of image quality evaluation model, image quality evaluation method and device |
CN115629930B (en) * | 2022-12-23 | 2023-08-08 | 北京东远润兴科技有限公司 | Fault detection method, device, equipment and storage medium based on DSP system |
CN116486273B (en) * | 2023-06-20 | 2023-09-01 | 南昌工程学院 | Method for extracting water body information of small sample remote sensing image |
-
2023
- 2023-08-24 CN CN202311076893.3A patent/CN116958122B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112102309A (en) * | 2020-09-27 | 2020-12-18 | 中国建设银行股份有限公司 | Method, device and equipment for determining image quality evaluation result |
CN112950581A (en) * | 2021-02-25 | 2021-06-11 | 北京金山云网络技术有限公司 | Quality evaluation method and device and electronic equipment |
Non-Patent Citations (1)
Title |
---|
"基于Curvelet变换的SAR图像增强";李林;《仪器仪表学报》;20060630;第27卷(第06期);第2134-2135页 * |
Also Published As
Publication number | Publication date |
---|---|
CN116958122A (en) | 2023-10-27 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN113498528B (en) | Image defect determining method and device, electronic equipment and storage medium | |
Singh et al. | A novel optical image denoising technique using convolutional neural network and anisotropic diffusion for real-time surveillance applications | |
CN116958122B (en) | SAR image evaluation method, SAR image evaluation device, SAR image evaluation equipment and readable storage medium | |
CN105654098B (en) | Hyperspectral remote sensing image classification method and system | |
CN103971127B (en) | Forward-looking radar imaging sea-surface target key point detection and recognition method | |
CN114820465A (en) | Point cloud detection model training method and device, electronic equipment and storage medium | |
CN110766007B (en) | Certificate shielding detection method, device, equipment and readable storage medium | |
CN111445058B (en) | Data analysis method, device, equipment and computer readable storage medium | |
Huang et al. | A new SAR image segmentation algorithm for the detection of target and shadow regions | |
CN112016569A (en) | Target detection method, network, device and storage medium based on attention mechanism | |
Serir et al. | No-reference blur image quality measure based on multiplicative multiresolution decomposition | |
WO2024139287A1 (en) | Road disease prediction method and apparatus, electronic device and storage medium | |
CN111144425A (en) | Method and device for detecting screen shot picture, electronic equipment and storage medium | |
CN110235177B (en) | Image processing device, image recognition device, and storage medium | |
Xie et al. | Reduced-reference image quality metric based on statistic model in complex wavelet transform domain | |
CN107657615A (en) | High Resolution SAR image change detection method based on increment CAE | |
Vitalii et al. | An automated method for finding the optimal parameters of adaptive filters for speckle denoising of SAR images | |
CN116030300A (en) | Progressive domain self-adaptive recognition method for zero-sample SAR target recognition | |
CN113962968B (en) | Multi-source mixed interference radar image target detection system oriented to complex electromagnetic environment | |
CN116109581A (en) | Concrete road monitoring method and device, electronic equipment and storage medium | |
CN115810105A (en) | Panorama segmentation method, device, equipment and storage medium | |
CN111127485B (en) | Method, device and equipment for extracting target area in CT image | |
Parse et al. | EDGE detection technique based on bilateral filtering and iterative threshold selection algorithm and transfer learning for traffic sign recognition | |
Khosravi et al. | A new paradigm for image quality assessment based on human abstract layers of quality perception | |
CN112712550A (en) | Image quality evaluation method and device |
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 |