CN115830351A - Image processing method, apparatus and storage medium - Google Patents
Image processing method, apparatus and storage medium Download PDFInfo
- Publication number
- CN115830351A CN115830351A CN202310115236.9A CN202310115236A CN115830351A CN 115830351 A CN115830351 A CN 115830351A CN 202310115236 A CN202310115236 A CN 202310115236A CN 115830351 A CN115830351 A CN 115830351A
- Authority
- CN
- China
- Prior art keywords
- image
- images
- similarity
- verified
- face
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Landscapes
- Image Analysis (AREA)
Abstract
The invention provides an image processing method, an image processing device and a storage medium, which belong to the technical field of image processing and specifically comprise the following steps: acquiring an image set to be processed and extracting images to obtain an image to be verified; the method comprises the steps that similarity evaluation is conducted on an image to be verified and other images of an image set to obtain similar images, when the number of the similar images is larger than a first number threshold value, the similar images and the image to be verified are used as the image set to be verified, and screening of optional images in the image set to be verified is achieved through image noise; evaluating the face similarity of the target face reference image and the selectable image, taking the image smaller than the threshold value as a candidate suspected unqualified image, and taking the rest images as secondary screening images; and obtaining an image quality score based on the image noise and the face similarity of the secondary screening image, taking the image quality score smaller than a threshold value as a candidate suspected unqualified image, and taking the rest images as recommended images, thereby further improving the screening efficiency of the high-quality images.
Description
Technical Field
The present invention relates to the field of image processing technologies, and in particular, to an image processing method, an image processing apparatus, and a storage medium.
Background
When image processing is carried out, particularly when a batch of figure photos are screened, a plurality of similar photos often exist, particularly when the number of the photos is large, screening of low-quality photos becomes a very difficult work, and in order to realize automatic screening of the photos, a human face picture test set is established in a method and a device for optimizing a human face picture quality evaluation model which are disclosed in an invention patent publication No. CN 107609493B; identifying the similarity between the face picture to be detected and a sample face picture in a preset face database, and obtaining the identification result of each face picture to be detected according to the similarity and the picture identity information; determining the quality score of each face picture to be detected according to the identification result; the method comprises the following steps of (1) carrying out neural network training by taking a face picture to be tested and a corresponding quality score thereof as training data to obtain an optimized face picture quality evaluation model and parameters, but has the following technical problems:
1. when the quality of an image is evaluated without considering image noise of the image and the like, and a low-quality image is identified, when the image is screened, particularly when the number of images is large, if the image with large image noise cannot be simultaneously screened first, the final processing efficiency is obviously reduced.
2. The determination of the number of similarities of the images is not considered, and when the number of similarities of some images is smaller than a certain threshold value, that is, the number of the same or similar images is small, the number of images that need to be distinguished by a person is already small, and at this time, screening is not needed, and if the number of similarities of the images cannot be determined, the final processing efficiency is also affected.
In view of the above technical problems, the present invention provides an image processing method, an apparatus, and a storage medium.
Disclosure of Invention
In order to realize the purpose of the invention, the invention adopts the following technical scheme:
according to an aspect of the present invention, there is provided an image processing method.
An image processing method is characterized by specifically comprising the following steps:
s11, acquiring a batch of image sets to be processed, and sequentially extracting images based on the image sets to obtain images to be verified;
s12, extracting image features based on the image to be verified, evaluating the similarity based on the image features and the image features of other images of the image set to obtain similar images with the similarity larger than a first threshold value, judging whether the number of the similar images is larger than a first number threshold value, if so, entering the step S13, otherwise, eliminating the similar images and the image to be verified, taking the image set after elimination as a new image set, and returning to the step S11;
s13, constructing an image set to be verified based on the similar image and the image to be verified, identifying image noise based on the image set to be verified, taking an image with the image noise larger than a first noise threshold value as a candidate suspected unqualified image, and taking the rest images in the image set to be verified as optional images;
s14, obtaining a target face reference image of the image set, obtaining the face similarity of the target face reference image and the selectable images by adopting a face similarity evaluation model based on a machine learning algorithm, taking the selectable images with the face similarity smaller than a similarity threshold value as candidate suspected unqualified images, and taking the remaining selectable images as secondary screening images;
s15, obtaining an image quality score based on the image noise and the face similarity of the secondary screening images, taking the secondary screening images with the image quality score smaller than the first quality threshold value as alternative suspected unqualified images, and taking the rest secondary screening images as recommended images.
Through the evaluation of the similar images, the technical problem that the alternative photos with few similar images or the alternative photos without similar images need too much screening is solved, the screening quantity of the suspected unqualified alternative images is further reduced, and the screening efficiency is improved.
The identification and screening of the alternative suspected unqualified images are realized based on the image noise, the face similarity and the image quality score, so that the number of images needing image screening is further reduced, the quality of the images is evaluated from multiple angles, the screening of high-quality images is realized, and the accuracy and the effectiveness of alternative photos are further improved.
In another aspect, an embodiment of the present application provides a computer device, including: a memory and a processor communicatively coupled, and a computer program stored on the memory and operable on the processor, wherein: an image processing method as described above when the processor runs the computer program.
In another aspect, the present invention provides a computer storage medium having a computer program stored thereon, which, when executed in a computer, causes the computer to perform one of the image processing methods described above.
Additional features and advantages will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and drawings.
In order to make the aforementioned and other objects, features and advantages of the present invention comprehensible, preferred embodiments accompanied with figures are described in detail below.
Drawings
The above and other features and advantages of the present invention will become more apparent by describing in detail exemplary embodiments thereof with reference to the attached drawings.
Fig. 1 is a flowchart of an image processing method according to embodiment 1;
FIG. 2 is a flowchart of the detailed steps of determining similar images with similarity greater than a first threshold to an image to be verified according to embodiment 1;
FIG. 3 is a flowchart of the detailed steps of face similarity determination according to example 1;
fig. 4 is a flowchart of specific steps of image quality score determination according to embodiment 1;
fig. 5 is a frame diagram of a computer storage medium in embodiment 3.
Detailed Description
Example embodiments will now be described more fully with reference to the accompanying drawings. Example embodiments may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein; rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the concept of example embodiments to those skilled in the art. The same reference numerals in the drawings denote the same or similar structures, and thus their detailed description will be omitted.
The terms "a", "an", "the", "said" are used to indicate the presence of one or more elements/components/etc.; the terms "comprising" and "having" are intended to be inclusive and mean that there may be additional elements/components/etc. other than the listed elements/components/etc.
The problems of the prior art are summarized as follows:
in some fixed occasions, such as wedding photo shooting, art photo shooting and the like, a large number of identical negative films are formed in the same posture, and the screening of the large number of identical negative films in a manual mode is often performed in the past, so that time and labor are consumed, and images with higher quality can be missed, and the need of solving the problem is high.
Example 1
In order to solve the above problem, according to an aspect of the present invention, as shown in fig. 1, there is provided an image processing method, specifically comprising:
s11, acquiring a batch of image sets to be processed, and sequentially extracting images based on the image sets to obtain images to be verified;
specifically, before extracting the images to be verified, the number of the images in the image set needs to be determined, and if and only if the number of the images in the image set is greater than a second number threshold, the images to be verified are extracted.
For example, if the number of images in the image set to be verified is less than or equal to 3, the images to be verified need not be extracted, and the amount of data to be processed is significantly low.
S12, extracting image features based on the image to be verified, evaluating the similarity based on the image features and the image features of other images of the image set to obtain similar images with the similarity larger than a first threshold value, judging whether the number of the similar images is larger than a first number threshold value, if so, entering a step S13, if not, excluding the similar images and the image to be verified, taking the image set after exclusion as a new image set, and returning to the step S11;
specifically, as shown in fig. 2, the specific step of determining the similar image whose similarity to the image to be verified is greater than the first threshold value is as follows:
s21, acquiring peak signal-to-noise ratios of the image to be verified and other images of the image set, taking the other images with the peak signal-to-noise ratios larger than a first signal-to-noise ratio threshold value as alternative similar images, and extracting features of a color histogram vector and a color moment vector based on the image to be verified to obtain a feature vector of the image to be verified;
in particular, peak signal-to-noise ratio (PSNR) is an engineering term that represents the ratio of the maximum possible power of a signal and the power of destructive noise that affects its representation accuracy. Since many signals have a very wide dynamic range, the peak signal-to-noise ratio is often expressed in logarithmic decibel units. The peak signal-to-noise ratio is often used as a measure of the quality of signal reconstruction in the field of image compression and the like, and is often defined simply by the Mean Square Error (MSE).
Specifically, the calculation formula of the peak signal-to-noise ratio is as follows:
specifically, the specific steps of extracting the color histogram vector are as follows:
1) Adjusting the size of the image, and normalizing the H component histogram obtained in HSV (Hue-Saturation-Value) space;
2) The value range of the H component in OpenCV is [0, 180], the H component is divided into 60 areas, and each area comprises 3 degree magnitudes;
3) Frequency numbers appearing in 3 degree orders of magnitude in each area are superposed and summed, and the frequencies of pixels in 60 areas are extracted to form a 60-dimensional vector v representing image characteristics 1 。
Specifically, the color moment vector extraction specifically comprises the following steps:
the color moment is a global feature for representing image color information, and because the color feature is mainly concentrated in a low-order moment, the color distribution of the image can be effectively represented by generally selecting a first-order moment, a second-order moment and a third-order moment.
The formula for the first, second and third moments of color of an image is expressed as follows:
wherein: p is a radical of ij The pixel value of the jth pixel point of the ith color channel component of the three-channel image is shown, and N is the number of pixels.
S22, extracting features of a color histogram vector and a color moment vector based on the alternative similar image to obtain a feature vector of the alternative similar image, matching feature points based on Euclidean distances between the feature vector of the image to be verified and the feature vector of the alternative similar image, and obtaining basic similarity between the image to be verified and the alternative similar image based on the feature points;
s23, calculating the structural similarity between the image to be verified and the alternative similar image, obtaining the similarity of the alternative similar image based on the peak signal-to-noise ratio, the basic similarity and the structural similarity of the alternative similar image, and taking the alternative similar image with the similarity larger than a first threshold value as the similar image with the similarity larger than the first threshold value with the image to be verified.
Specifically, the first threshold is determined according to the number of images in the image set, wherein the larger the number of images in the image set is, the larger the first threshold is.
Specifically, the calculation formula of the similarity is as follows:
wherein min is a minimum function, Z is a peak signal-to-noise ratio of the alternative similar image, and S1 and S2 are respectively a basic similarity and a structural similarity.
Specifically, for example, when the similarity of the candidate similar image is 0.69 and the first threshold is 0.6, the candidate similar image is regarded as the similar image.
Specifically, the similar images and the images to be verified are excluded, and the excluded image set is used as a new image set, so that the exclusion with less image similarity number of the images to be verified is realized.
S13, constructing an image set to be verified based on the similar image and the image to be verified, identifying image noise based on the image set to be verified, taking an image with the image noise larger than a first noise threshold value as a candidate suspected unqualified image, and taking the rest images in the image set to be verified as optional images;
specifically, for example, the identification of the image noise is to perform filtering processing on the image set to be verified in a gaussian filtering manner to obtain a filtered image set, and obtain the image noise of the image set to be verified based on a difference between the filtered image set and the image set to be verified.
Specifically, for example, the key code for determining the image noise is:
the key code of the algorithm process is realized as follows:
- (void)filter
{
[MBProgressHUD showHUDAddedTo:self.view animated:YES];
v/callback after completion of judgment
NSOperation *completeOperation = [NSBlockOperation blockOperationWithBlock:^
{
dispatch_async(dispatch_get_main_queue(), ^{
// judging similar images
NSMutableArray *photoPool = [self.photoAssets mutableCopy];
[self comparePhotos:photoPool complete:^{
[MBProgressHUD hideHUDForView:self.view animated:YES];
[self.collectionView reloadData];
}];
});
}];
// determination of blur, exposure, etc
for (PhotoModel *model in self.photoAssets) {
AnalyzeOperation *op = [[AnalyzeOperation alloc] initWithModel:model];
[self.queue addOperation:op];
[completeOperation addDependency:op];}
[self.queue addOperation:completeOperation];
}
Specifically, besides image noise, the degree of blur of the similar image may be identified, and an image with the degree of blur larger than a certain threshold may be used as a candidate suspected unqualified image.
Specifically, the objective evaluation may be classified into: full Reference Image Blur evaluation (FR-IBA), partial Reference Image Blur evaluation (RR-IBA), and No Reference Image Blur evaluation (NR-IBA). The objective ambiguity evaluation method can also refer to an objective image quality evaluation method, but only focuses on one index of ambiguity, so that the method is more targeted in algorithm design, and the emphasis should be placed on the extraction of the ambiguity characteristic parameters.
The ambiguity assessment algorithm can be in any of the following categories: (1) Pixel-based techniques, including analyzing statistical properties of pixel gray-scale values and correlations between pixels; (2) The technology based on the transform domain utilizes the principle that the more high-frequency components in the transform domain, the clearer the image is, and the less high-frequency components, the more fuzzy the image is; (3) The image gradient-based technology measures the blurring degree of an image by utilizing the gradient of the edge of the image, and the image is clearer when the gradient is larger.
Wherein the calculation formula of the ambiguity is as follows:
where f (x, y) $ is the original image, d (x, y) $ is the Point Spread Function (PSF),
is convolution and n (x, y) is additive noise.
S14, obtaining a target face reference image of the image set, obtaining the face similarity of the target face reference image and the selectable images by adopting a face similarity evaluation model based on a machine learning algorithm, taking the selectable images with the face similarity smaller than a similarity threshold value as candidate suspected unqualified images, and taking the remaining selectable images as secondary screening images;
specifically, as shown in fig. 3, the specific steps of determining the similarity between human faces are as follows:
s31, based on the selectable images, adopting a face detection model based on a DNN neural network to detect image areas in the selectable images with confidence degrees larger than a first confidence degree threshold value as face images;
s32, based on the face image and the target face reference image, carrying out LBP feature extraction to respectively obtain LBP features of the face image and the target face reference image;
s33, the face image and the target face reference image are uniformly divided into a pair of sub-regions, histograms of the sub-regions are counted according to LBP values, the histograms are used as distinguishing features of the sub-regions, and the face similarity between the face image and the target face reference image is determined by adopting a cosine similarity method.
Specifically, the target face reference image of the image set is determined according to a historical face image of a target person of the image set.
S15, obtaining an image quality score based on the image noise and the face similarity of the secondary screening images, taking the secondary screening images with the image quality scores smaller than the first quality threshold value as alternative suspected unqualified images, and taking the rest secondary screening images as recommended images.
Specifically, as shown in fig. 4, the specific steps of determining the image quality score are as follows:
s41, constructing an input set based on the image noise and the face similarity of the secondary screening image;
s42, transmitting the input set to an image quality evaluation model based on an SSA-PNN neural network algorithm to obtain an evaluation result;
specifically, for example, the image quality assessment model is constructed by the following specific steps:
step 1: SSA algorithm parameters are initialized. The specific parameters include initial sparrow number n, initial sparrow positions, ratio of discoverers to joiners, upper limit value of iteration times, upper and lower limit values, population warning value R2, safety value ST and other parameters.
Step 2: and calculating the fitness value of each virtual sparrow.
And step 3: and sequentially updating the positions of the discoverer, the joiner and the alerter in the virtual sparrow group.
And 4, step 4: and (5) circularly iterating until the circularly iteration times are set initially, and if the iteration times are not full, turning to the step 2.
And 5: and substituting the smoothing factor parameters obtained by SSA optimization into the PNN, starting network training on the training group data by the PNN, judging the input data by the PNN after training is finished, and obtaining the image quality scoring result of the model.
Specifically, the value range of the image quality score is between 0 and 1, wherein the larger the image quality score is, the higher the image quality of the image is.
Specifically, the location update formula of the discoverer is as follows:
specifically, after the position of the optimal individual of the population is updated, cauchy variation disturbance is added to the population, the fitness value of the disturbed sparrows is calculated and compared with that of the disturbed sparrows, the sparrow population with low fitness value is reserved, and then the position of the optimal individual of the current population is recorded;
specifically, the Cauchy variation is obtained from Cauchy distribution of continuous probability, and is mainly characterized in that the peak value at zero is small, and the peak value is slowly reduced to zero, so that the variation range is more uniform.
In the formula: x is the original individual position of the cell,
u is a random number between 0 and 1, which is the position of an individual after Cauchy mutation.
S43 determining an image quality score of the secondary screening image based on the evaluation result.
Specifically, the candidate suspected unqualified image and the recommended image are stored by adopting different folders respectively.
Specifically, for example, the suspected rejected candidate image is placed in the suspected rejected candidate image folder, and the recommended image is placed in the recommended image folder.
For example, as shown in table 1, 97 negative films are generated in total in the actual shooting process, and correspond to 10 different photos, respectively, wherein the number of generated recommended images is 29, and the number of generated unqualified images is 68, so that the image screening efficiency is greatly improved, and the specific results of image screening and recommended image output by using the method are as follows:
TABLE 1 recommendation image Generation results
| Type of photograph | Number of similar images | Number of non-conforming images | Recommending number of images |
| Photograph 1 | 12 | 8 | 4 |
| Photograph 2 | 15 | 9 | 6 |
| Photograph 3 | 9 | 7 | 2 |
| Photograph 4 | 11 | 8 | 3 |
| Photograph 5 | 15 | 10 | 5 |
| Photograph 6 | 18 | 12 | 3 |
| Photograph 7 | 17 | 11 | 6 |
Through the evaluation of the similar images, the technical problem that the alternative photos with few similar images or the alternative photos without similar images need too much screening is solved, the screening quantity of the suspected unqualified alternative images is further reduced, and the screening efficiency is improved.
The identification and screening of the alternative suspected unqualified images are realized based on the image noise, the face similarity and the image quality score, so that the number of images needing image screening is further reduced, the quality of the images is evaluated from multiple angles, the screening of high-quality images is realized, and the accuracy and the effectiveness of alternative photos are further improved.
Example 2
An embodiment of the present application provides a computer device, including: a memory and a processor communicatively coupled, and a computer program stored on the memory and operable on the processor, wherein: an image processing method as described above when the processor runs the computer program.
Example 3
As shown in fig. 5, the present invention provides a computer storage medium having a computer program stored thereon, which, when executed in a computer, causes the computer to execute one of the image processing methods described above.
In the several embodiments provided in the present application, it should be understood that the disclosed system and method may be implemented in other manners. The system embodiments described above are merely illustrative, and for example, the flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present invention. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
In addition, the functional modules in the embodiments of the present invention may be integrated together to form an independent part, or each module may exist separately, or two or more modules may be integrated to form an independent part.
The functions, if implemented in the form of software functional modules and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
In light of the foregoing description of the preferred embodiment of the present invention, many modifications and variations will be apparent to those skilled in the art without departing from the spirit and scope of the invention. The technical scope of the present invention is not limited to the content of the specification, and must be determined according to the scope of the claims.
Claims (12)
1. An image processing method is characterized by specifically comprising the following steps:
s11, acquiring a batch of image sets to be processed, and sequentially extracting images based on the image sets to obtain images to be verified;
s12, extracting image features based on the image to be verified, evaluating the similarity based on the image features and the image features of other images of the image set to obtain similar images with the similarity larger than a first threshold value, judging whether the number of the similar images is larger than a first number threshold value, if so, entering a step S13, if not, excluding the similar images and the image to be verified, taking the image set after exclusion as a new image set, and returning to the step S11;
s13, constructing an image set to be verified based on the similar image and the image to be verified, identifying image noise based on the image set to be verified, taking the image with the image noise larger than a first noise threshold value as a candidate suspected unqualified image, and taking the rest images in the image set to be verified as selectable images;
s14, obtaining a target face reference image of the image set, obtaining the face similarity of the target face reference image and the selectable image by adopting a face similarity evaluation model based on a machine learning algorithm, taking the selectable image with the face similarity smaller than a similarity threshold value as a suspected candidate failing image, and taking the rest selectable images as secondary screening images;
s15, obtaining an image quality score based on the image noise and the face similarity of the secondary screening images, taking the secondary screening images with the image quality scores smaller than the first quality threshold value as alternative suspected unqualified images, and taking the rest secondary screening images as recommended images.
2. The image processing method according to claim 1, wherein before the extraction of the image to be verified, the number of images in the image set is determined, and if and only if the number of images in the image set is greater than the second number threshold, the extraction of the image to be verified is performed.
3. The image processing method according to claim 1, wherein the specific step of determining the similar image whose similarity with the image to be verified is greater than the first threshold value is:
acquiring peak signal-to-noise ratios of the image to be verified and other images of the image set, taking the other images with the peak signal-to-noise ratios larger than a first signal-to-noise ratio threshold value as alternative similar images, and performing feature extraction of color histogram vectors and color moment vectors on the basis of the image to be verified to obtain feature vectors of the image to be verified;
extracting the features of a color histogram vector and a color moment vector based on the alternative similar image to obtain a feature vector of the alternative similar image, matching feature points based on Euclidean distances between the feature vector of the image to be verified and the feature vector of the alternative similar image, and obtaining the basic similarity between the image to be verified and the alternative similar image based on the feature points;
and calculating the structural similarity between the image to be verified and the alternative similar image, obtaining the similarity of the alternative similar image based on the peak signal-to-noise ratio, the basic similarity and the structural similarity of the alternative similar image, and taking the alternative similar image with the similarity larger than a first threshold value as the similar image with the similarity larger than the first threshold value with the image to be verified.
4. An image processing method as claimed in claim 3, characterized in that the color moment vectors are represented by third-order moments.
5. An image processing method as claimed in claim 1, characterized in that the first threshold value is determined on the basis of the number of images of the image set, wherein the greater the number of images of the image set, the greater the first threshold value.
6. An image processing method as claimed in claim 4, characterized in that the similarity is calculated by the formula:
7. The image processing method according to claim 1, wherein the step of determining the similarity between human faces comprises:
based on the selectable image, adopting a human face detection model based on a DNN neural network to detect an image area in the selectable image with the confidence coefficient larger than a first confidence coefficient threshold value as a human face image;
carrying out LBP characteristic extraction based on the face image and the target face reference image to respectively obtain LBP characteristics of the face image and the target face reference image;
and uniformly dividing the face image and the target face reference image into a pair of sub-regions, counting histograms of the sub-regions according to LBP values, taking the histograms as distinguishing features of the sub-regions, and determining the face similarity of the face image and the target face reference image by adopting a cosine similarity method.
8. An image processing method as claimed in claim 1, characterized in that the target face reference image of the image set is determined from historical face images of the target person of the image set.
9. The image processing method as claimed in claim 1, wherein the image quality score is determined by the specific steps of:
constructing an input set based on the image noise and the face similarity of the secondary screening image;
transmitting the input set to an image quality evaluation model based on an SSA-PNN neural network algorithm to obtain an evaluation result;
determining an image quality score for the secondary screening image based on the evaluation result.
10. The image processing method according to claim 1, wherein the candidate suspected-to-be-failed image and the recommended image are stored in different folders, respectively.
11. A computer device, comprising: a memory and a processor communicatively coupled, and a computer program stored on the memory and operable on the processor, wherein: the processor, when executing the computer program, performs an image processing method according to any of claims 1-10.
12. A computer storage medium having stored thereon a computer program which, when executed in a computer, causes the computer to execute an image processing method according to any one of claims 1 to 10.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310115236.9A CN115830351B (en) | 2023-02-15 | 2023-02-15 | Image processing method, apparatus and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310115236.9A CN115830351B (en) | 2023-02-15 | 2023-02-15 | Image processing method, apparatus and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115830351A true CN115830351A (en) | 2023-03-21 |
| CN115830351B CN115830351B (en) | 2023-04-28 |
Family
ID=85521457
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310115236.9A Active CN115830351B (en) | 2023-02-15 | 2023-02-15 | Image processing method, apparatus and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115830351B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116055900A (en) * | 2023-03-30 | 2023-05-02 | 北京城建智控科技股份有限公司 | Image quality correction method based on image pickup device |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108269254A (en) * | 2018-01-17 | 2018-07-10 | 百度在线网络技术（北京）有限公司 | Image quality measure method and apparatus |
| CN110276284A (en) * | 2019-06-11 | 2019-09-24 | 暨南大学 | Flame identification method, device, equipment and storage medium based on video quality assessment |
| CN110929545A (en) * | 2018-09-19 | 2020-03-27 | 传线网络科技（上海）有限公司 | Human face image sorting method and device |
| CN113505854A (en) * | 2021-07-29 | 2021-10-15 | 济南博观智能科技有限公司 | Method, device, equipment and medium for constructing facial image quality evaluation model |
| CN113706502A (en) * | 2021-08-26 | 2021-11-26 | 重庆紫光华山智安科技有限公司 | Method and device for evaluating quality of face image |
| WO2022100337A1 (en) * | 2020-11-11 | 2022-05-19 | 腾讯科技（深圳）有限公司 | Face image quality assessment method and apparatus, computer device and storage medium |
-
2023
- 2023-02-15 CN CN202310115236.9A patent/CN115830351B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108269254A (en) * | 2018-01-17 | 2018-07-10 | 百度在线网络技术（北京）有限公司 | Image quality measure method and apparatus |
| CN110929545A (en) * | 2018-09-19 | 2020-03-27 | 传线网络科技（上海）有限公司 | Human face image sorting method and device |
| CN110276284A (en) * | 2019-06-11 | 2019-09-24 | 暨南大学 | Flame identification method, device, equipment and storage medium based on video quality assessment |
| WO2022100337A1 (en) * | 2020-11-11 | 2022-05-19 | 腾讯科技（深圳）有限公司 | Face image quality assessment method and apparatus, computer device and storage medium |
| CN113505854A (en) * | 2021-07-29 | 2021-10-15 | 济南博观智能科技有限公司 | Method, device, equipment and medium for constructing facial image quality evaluation model |
| CN113706502A (en) * | 2021-08-26 | 2021-11-26 | 重庆紫光华山智安科技有限公司 | Method and device for evaluating quality of face image |
Non-Patent Citations (2)
| Title |
|---|
| JIAO DU等: "An overview of multi-modal medical image fusion" * |
| 陈中钱;: "监控视频图像质量无参考型客观评价方法研究" * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116055900A (en) * | 2023-03-30 | 2023-05-02 | 北京城建智控科技股份有限公司 | Image quality correction method based on image pickup device |
| CN116055900B (en) * | 2023-03-30 | 2023-06-09 | 北京城建智控科技股份有限公司 | Image quality correction method based on image pickup device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115830351B (en) | 2023-04-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107133948B (en) | Image blurring and noise evaluation method based on multitask convolution neural network | |
| CN108765465B (en) | Unsupervised SAR image change detection method | |
| CN112215201B (en) | Method and device for evaluating face recognition model and classification model aiming at image | |
| CN112800876B (en) | Super-spherical feature embedding method and system for re-identification | |
| Cord et al. | Texture classification by statistical learning from morphological image processing: application to metallic surfaces | |
| CN104408707A (en) | Rapid digital imaging fuzzy identification and restored image quality assessment method | |
| US9245198B2 (en) | Object recognition by comparison of patterns against map of image | |
| CN111753873A (en) | Image detection method and device | |
| CN112634171B (en) | Image defogging method and storage medium based on Bayesian convolutional neural network | |
| CN116740728B (en) | Dynamic acquisition method and system for wafer code reader | |
| CN108830829B (en) | Non-reference quality evaluation algorithm combining multiple edge detection operators | |
| CN115830351B (en) | Image processing method, apparatus and storage medium | |
| CN112926429A (en) | Machine audit model training method, video machine audit method, device, equipment and storage medium | |
| CN111814852A (en) | Image detection method, image detection device, electronic equipment and computer-readable storage medium | |
| CN112528777A (en) | Student facial expression recognition method and system used in classroom environment | |
| CN109376782B (en) | Support vector machine cataract classification method and device based on eye image features | |
| Kanchev et al. | Blurred image regions detection using wavelet-based histograms and SVM | |
| Alaql et al. | Classification of image distortions for image quality assessment | |
| CN115984178A (en) | Counterfeit image detection method, electronic device, and computer-readable storage medium | |
| Alsandi | Image splicing detection scheme using surf and mean-LBP based morphological operations | |
| CN112257688A (en) | GWO-OSELM-based non-contact palm in-vivo detection method and device | |
| Wang et al. | Median filtering detection using LBP encoding pattern★ | |
| CN117011196B (en) | Infrared small target detection method and system based on combined filtering optimization | |
| CN117474916B (en) | Image detection method, electronic equipment and storage medium | |
| Kaur et al. | A novel approach to no-reference image quality assessment using canny magnitude based upon neural network |
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 |