CN115760609A - Image optimization method and system - Google Patents
Image optimization method and system Download PDFInfo
- Publication number
- CN115760609A CN115760609A CN202211417961.3A CN202211417961A CN115760609A CN 115760609 A CN115760609 A CN 115760609A CN 202211417961 A CN202211417961 A CN 202211417961A CN 115760609 A CN115760609 A CN 115760609A
- Authority
- CN
- China
- Prior art keywords
- image
- optimization
- similarity
- mode
- optimized
- 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
- 238000005457 optimization Methods 0.000 title claims abstract description 144
- 238000000034 method Methods 0.000 title claims abstract description 40
- 239000011159 matrix material Substances 0.000 claims description 24
- 230000009466 transformation Effects 0.000 claims description 24
- 238000004364 calculation method Methods 0.000 claims description 18
- 238000012549 training Methods 0.000 claims description 12
- 238000012545 processing Methods 0.000 claims description 9
- 238000012360 testing method Methods 0.000 claims description 9
- 230000002146 bilateral effect Effects 0.000 claims description 6
- 238000001914 filtration Methods 0.000 claims description 6
- 238000005286 illumination Methods 0.000 claims description 5
- 238000013528 artificial neural network Methods 0.000 claims description 3
- 230000008569 process Effects 0.000 description 3
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000011218 segmentation Effects 0.000 description 2
- 230000002776 aggregation Effects 0.000 description 1
- 238000004220 aggregation Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Image Analysis (AREA)
- Facsimile Image Signal Circuits (AREA)
Abstract
The invention discloses an image optimization method and system, firstly, inputting an image to be optimized into an image optimal model for optimization to obtain an initial optimized image; secondly, calculating the similarity between the initial optimized image and the standard image; then judging whether the similarity belongs to a fifth set threshold range or not; if the image belongs to the range of the fifth set threshold, taking the initial optimized image as a final optimized image; if the similarity does not belong to the range of the fifth set threshold, selecting a corresponding optimization mode according to the similarity; optimizing the initial optimized image by using the selected optimization mode; and finally, taking the optimized image as an initial optimized image, and returning to recalculate the similarity between the initial optimized image and the standard image. The technical scheme disclosed by the invention realizes repeated optimization of the image, and only when the similarity belongs to the range of the fifth set threshold, the final optimized image is output, so that the quality of image output is improved.
Description
Technical Field
The invention belongs to the technical field of image optimization, and particularly relates to an image optimization method and system.
Background
With the development of the camera technology, the performance of various camera terminals is better and better at present, and the requirements of people on images are higher and higher. Because the situations of the definition, the noise point and the like of the images acquired by the different types of camera terminals are different, and the acquired images are optimized by only using a single image optimization model, the problems that the optimized images still have low pixels, the images are not clear and the like are likely to exist, so that the technical staff in the field needs to set an optimization mode urgently to enable the optimized images to reach the optimal state and improve the quality of image output.
Disclosure of Invention
Based on the above technical problem, the present invention provides an image optimization method and system to improve the quality of image output.
The specific technical scheme is as follows:
the invention discloses an image optimization method, which comprises the following steps:
acquiring an image to be optimized and a standard image;
inputting the image to be optimized into an image optimal model for optimization to obtain an initial optimized image;
calculating the similarity between the initial optimized image and the standard image;
judging whether the similarity belongs to a fifth set threshold range or not; if the image belongs to a fifth set threshold range, taking the initial optimized image as a final optimized image; if the similarity does not belong to the range of a fifth set threshold, selecting a corresponding optimization mode according to the similarity;
optimizing the initial optimized image by using the selected optimization mode;
and taking the optimized image as an initial optimized image, and returning to the step of calculating the similarity between the initial optimized image and the standard image.
In the above scheme of the image optimization method, before the step of "obtaining the image to be optimized and the standard image", the method further includes:
acquiring a sample image set obtained by acquiring target objects by a plurality of different types of image collectors and an annotation image set corresponding to the sample image;
selecting a training set from the sample image set and the labeled image set according to a first set proportion; selecting a test set from the rest of the sample image sets and the rest of the labeled image sets according to a second set proportion;
training a neural network by using the training set to obtain an image initial optimization model;
verifying the initial image optimization model by using the test set to obtain accuracy;
judging whether the accuracy reaches a set value, and if the accuracy is greater than or equal to the set value, taking the image initial optimization model as an image optimal model; and if the accuracy is smaller than the set value, reselecting the training set and the test set.
In the above scheme of the image optimization method, the calculating a similarity between the initial optimized image and the standard image includes:
dividing a target identification area on the initial optimization image to obtain a first alignment area;
dividing a target identification area on the standard image to obtain a second alignment area;
carrying out format unification processing on the first alignment area and the second alignment area to obtain a third alignment area and a fourth alignment area; the third alignment region and the fourth alignment region both comprise mxn pixel points; wherein m and n are both positive integers greater than 5;
and calculating the similarity according to each pixel point in the third alignment area and each pixel point in the fourth alignment area.
In the above scheme of the image optimization method, the calculating the similarity according to each pixel point in the third alignment region and each pixel point in the fourth alignment region includes:
calculating a mean value according to the pixel mean value corresponding to the fourth alignment area;
setting all the pixel points which are larger than or equal to the mean value as 1, and setting all the pixel points which are smaller than the mean value as 0, and obtaining a first transformation matrix corresponding to the third alignment area and a second transformation matrix corresponding to the fourth alignment area;
and calculating the similarity according to the first transformation matrix and the second transformation matrix.
In the above scheme of the image optimization method, selecting a corresponding optimization mode according to the similarity specifically includes:
judging whether the similarity belongs to a first set threshold range or not; if the data belong to the range of the first set threshold, adopting a first mode as an optimization mode; the first mode is to perform illumination enhancement operation on an image;
judging whether the similarity belongs to a second set threshold range; if the current time is within the range of a second set threshold, adopting a mode two as an optimization mode; the second mode is bilateral filtering and correcting operation;
judging whether the similarity belongs to a third set threshold range or not; if the current time is within the range of a third set threshold, adopting a mode three as an optimization mode; the third mode is to perform exposure enhancement operation on the image;
judging whether the similarity belongs to a fourth set threshold range or not; if the current value belongs to the range of the fourth set threshold, adopting a mode four as an optimization mode; and the fourth mode is to perform stretching operation on the saturation component.
In the above scheme of the image optimization method, after the optimizing the initial optimized image by using the selected optimization mode, the method further includes:
judging whether the optimization times reach an optimization time threshold value or not; if the optimization times reach the threshold of the optimization times, taking the initial optimization image as a final optimization image; and if the optimization times do not reach the threshold value of the optimization times, executing the step of taking the optimized image as the initial optimized image.
The invention also discloses an image optimization system, which comprises:
the acquisition module is used for acquiring an image to be optimized and a standard image;
the first optimization module is used for inputting the image to be optimized into an image optimal model for optimization to obtain an initial optimized image;
the similarity calculation module is used for calculating the similarity between the initial optimized image and the standard image;
the judging module is used for judging whether the similarity belongs to a fifth set threshold range; if the image belongs to the range of a fifth set threshold, taking the initial optimized image as a final optimized image; if the current time does not fall into the range of the fifth set threshold value, executing an optimization mode selection module;
the optimization mode selection module is used for selecting a corresponding optimization mode according to the similarity;
the second optimization module is used for optimizing the initial optimization image by using the selected optimization mode;
and the returning module is used for taking the optimized image as an initial optimized image and returning to the similarity calculating module.
In the above scheme of the image optimization system, the similarity calculation module specifically includes:
a first area dividing unit, configured to perform target identification area division on the initial optimized image to obtain a first alignment area;
the second area dividing unit is used for dividing the target identification area on the standard image to obtain a second alignment area;
the format unification processing unit is used for carrying out format unification processing on the first alignment area and the second alignment area to obtain a third alignment area and a fourth alignment area; the third alignment region and the fourth alignment region both comprise mxn pixel points; wherein m and n are both positive integers greater than 5;
and the similarity calculation unit is used for calculating the similarity according to each pixel point in the third alignment area and each pixel point in the fourth alignment area.
In the foregoing scheme of the image optimization system, the similarity calculation unit specifically includes:
the average value calculating subunit is configured to calculate an average value according to the pixel average value corresponding to the fourth alignment area;
the assignment subunit is configured to set, to 1, each pixel point that is greater than or equal to the mean value, and set, to 0, each pixel point that is smaller than the mean value, so as to obtain a first transformation matrix corresponding to the third alignment area and a second transformation matrix corresponding to the fourth alignment area;
and the similarity calculation subunit is used for calculating the similarity according to the first transformation matrix and the second transformation matrix.
In the above scheme of the image optimization system, the optimization mode selection module specifically includes:
the first judging unit is used for judging whether the similarity belongs to a first set threshold range or not; if the current time is within the first set threshold range, adopting a first mode as an optimization mode; the first mode is to perform illumination enhancement operation on an image;
the second judging unit is used for judging whether the similarity belongs to a second set threshold range or not; if the current time is within the range of a second set threshold, adopting a mode two as an optimization mode; the second mode is bilateral filtering and correcting operation;
the third judging unit is used for judging whether the similarity belongs to a third set threshold range or not; if the current time is within the range of a third set threshold, adopting a mode three as an optimization mode; the third mode is to perform exposure enhancement operation on the image;
a fourth judging unit, configured to judge whether the similarity falls within a fourth set threshold range; if the current time is within the range of the fourth set threshold, adopting a mode four as an optimization mode; and the fourth mode is to perform stretching operation on the saturation component.
Compared with the prior art, the invention discloses the following beneficial effects:
the invention discloses an image optimization method and system, firstly, inputting an image to be optimized into an image optimal model for optimization to obtain an initial optimized image; secondly, calculating the similarity between the initial optimized image and the standard image; then judging whether the similarity belongs to a fifth set threshold range or not; if the image belongs to the range of the fifth set threshold, taking the initial optimized image as a final optimized image; if the similarity does not belong to the range of the fifth set threshold, selecting a corresponding optimization mode according to the similarity; optimizing the initial optimized image by using the selected optimization mode; and finally, taking the optimized image as an initial optimized image, and returning to recalculate the similarity between the initial optimized image and the standard image. The technical scheme disclosed by the invention realizes repeated optimization of the image, and only when the similarity belongs to the range of the fifth set threshold, the final optimized image is output, so that the quality of image output is improved.
Drawings
FIG. 1 is a flow chart of an image optimization method of the present invention;
fig. 2 is a structural diagram of an image optimization system according to the present invention.
Detailed Description
The invention will be further described with reference to specific embodiments and drawings, but the invention is not limited to these embodiments.
Example 1
As shown in fig. 1, the present invention discloses an image optimization method, which comprises:
step S1: and acquiring an image to be optimized and a standard image.
Step S2: and inputting the image to be optimized into the image optimal model for optimization to obtain an initial optimized image.
And step S3: and calculating the similarity between the initial optimized image and the standard image.
And step S4: judging whether the similarity belongs to a fifth set threshold range or not; if the image belongs to the range of the fifth set threshold, taking the initial optimized image as a final optimized image; and if the similarity does not belong to the range of the fifth set threshold, selecting a corresponding optimization mode according to the similarity.
Step S5: and optimizing the initial optimization image by using the selected optimization mode.
Step S6: and taking the optimized image as an initial optimized image, and returning to the step S3.
The following is a detailed discussion of the various steps:
step S1: and acquiring an image to be optimized and a standard image.
The standard image and the image to be optimized are in one-to-one correspondence, the standard image is an image with annotation information, and the annotation information of the image comprises data information or image information which should be embodied by the image, such as object contour lines of the image, image results of artificial judgment and other annotation information. In addition, the image to be optimized can be the image acquired by the camera terminals of different types, and can also be the image acquired by the camera terminal of the same type, so that the process that the image acquired by the camera terminal of the same type can be optimized in the prior art is avoided.
As an optional implementation manner, before the step of "acquiring the image to be optimized and the standard image", the method further includes:
step S7: determining an image optimal model, and the specific steps comprise:
step S71: acquiring a plurality of different types of image collectors to acquire a target object to obtain a sample image set and an annotation image set corresponding to the sample image.
In the embodiment, the same type of image collector collects a plurality of different types of sample images, so that the constructed sample image set is more representative, the accuracy of constructing an image optimal model is further improved, and various types of images to be optimized can be identified and optimized. In addition, the image collector mentioned here may be a camera, a mobile phone, a tablet, a recorder, etc., and the above is only an example, as long as the image collection is realized.
Step S72: selecting a training set from the sample image set and the labeled image set according to a first set proportion; and selecting a test set from the rest of the sample image sets and the rest of the marked image sets according to a second set proportion.
In this embodiment, the first setting ratio and the second setting ratio are preferably set to 70% and 30%, and the specific setting ratio may also be selected according to actual requirements, which are not discussed one by one here.
Step S73: and training the neural network by using a training set to obtain an image initial optimization model.
Step S74: and verifying the initial image optimization model by using the test set to obtain the accuracy.
Step S75: judging whether the accuracy reaches a set value, and if the accuracy is greater than or equal to the set value, taking the image initial optimization model as an image optimal model; if the accuracy is smaller than the set value, the training set and the test set are reselected.
And step S3: calculating the similarity between the initial optimized image and the standard image, and specifically comprising the following steps:
step S31: and dividing a target identification area on the initial optimization image to obtain a first alignment area.
Step S32: and dividing a target identification area on the standard image to obtain a second alignment area.
In this embodiment, the first alignment area and the second alignment area may be obtained by performing area division on the initialization image and the standard image respectively by using a target identification method, but since the target size and the target offset are different in the target identification process, the two obtained alignment areas (i.e., the first alignment area and the second alignment area) may have different sizes, and therefore the two alignment areas need to be changed into a uniform specification, so as to facilitate subsequent similarity calculation.
Step S33: carrying out format unification processing on the first alignment area and the second alignment area to obtain a third alignment area and a fourth alignment area; the third alignment area and the fourth alignment area both comprise mxn pixel points; wherein m and n are both positive integers greater than 5, and the values of m and n can be selected according to actual requirements.
The format unification mentioned in this embodiment may be a scaling process, and may also be a region segmentation method, and the region segmentation method mentioned herein may include region growing, region separation and aggregation, a watershed method, and the like, which are only given as an example and will not be discussed more.
Step S34: and calculating the similarity according to each pixel point in the third alignment area and each pixel point in the fourth alignment area.
The similarity is calculated by two methods, wherein the first method adopts a mean value method and calculates the similarity according to each pixel point in a third alignment area and each pixel point in a fourth alignment area, and the second method adopts a gradient method and calculates the similarity according to each pixel point in the third alignment area and each pixel point in the fourth alignment area.
And calculating the similarity according to each pixel point in the third alignment area and each pixel point in the fourth alignment area by adopting a mean value method, wherein the method comprises the following specific steps:
step S341: and calculating the mean value according to the pixel mean value corresponding to the fourth alignment area.
Step S342: and setting the pixel points which are larger than or equal to the mean value as 1, and setting the pixel points which are smaller than the mean value as 0, so as to obtain a first transformation matrix corresponding to the third alignment area and a second transformation matrix corresponding to the fourth alignment area.
Step S343: and calculating the similarity according to the first transformation matrix and the second transformation matrix.
And calculating the similarity according to each pixel point in the third alignment area and each pixel point in the fourth alignment area by adopting a gradient method, wherein a specific calculation formula is as follows:
wherein S is p,q The similarity of a third alignment region p and a fourth alignment region q is represented, mxn represents the total number of pixel points in each alignment region, i, j respectively represent the coordinates of the pixel points in the third alignment region p, k, l respectively represent the coordinates of the pixel points in the fourth alignment region q, and c is a constant.
Step S5: selecting a corresponding optimization mode according to the similarity, which specifically comprises the following steps:
step S51: judging whether the similarity belongs to a first set threshold range; if the current time is within the first set threshold range, adopting a first mode as an optimization mode; if the current time does not fall within the first set threshold range, executing step S52; the first mode is to perform a lighting enhancement operation on the image.
Step S52: judging whether the similarity belongs to a second set threshold range; if the current value belongs to the range of the second set threshold, adopting a mode two as an optimization mode; if the current value does not fall within the second set threshold value range, executing step S53; mode two is bilateral filtering and calibration operations.
Step S53: judging whether the similarity belongs to a third set threshold range; if the current time is within the range of a third set threshold, adopting a mode three as an optimization mode; if the current time does not fall within the third set threshold range, executing step S54; and the third mode is to perform exposure enhancement operation on the image.
Step S54: judging whether the similarity belongs to a fourth set threshold range; if the current time is within the range of the fourth set threshold, adopting a mode four as an optimization mode; and the fourth mode is to perform stretching operation on the saturation component.
In this embodiment, the similarity value is in a range of 0 to 1, the first set threshold range may be 0 to 0.2, the second set threshold range may be 0.2 to 0.4, the third set threshold range may be 0.4 to 0.6, the fourth set threshold range may be 0.6 to 0.8, and the fifth set threshold range may be 0.8 to 1. The specific range can be selected according to actual requirements, and the above is only an example.
After the step of optimizing the initial optimized image by using the selected optimization mode, the method further comprises a step S8:
judging whether the optimization times reach an optimization time threshold value; if the optimization times reach the threshold value of the optimization times, taking the initial optimization image as a final optimization image; if the optimization number does not reach the optimization number threshold, "step S6" is performed.
Example 2
As shown in fig. 2, the present invention discloses an image optimization system, which comprises:
the acquisition module 1 is used for acquiring an image to be optimized and a standard image.
And the first optimization module 2 is used for inputting the image to be optimized into an image optimal model for optimization to obtain an initial optimized image.
And the similarity calculation module 3 is used for calculating the similarity between the initial optimized image and the standard image.
The judging module 4 is used for judging whether the similarity belongs to a fifth set threshold range; if the image belongs to the range of a fifth set threshold, taking the initial optimized image as a final optimized image; if the current time does not fall within the range of the fifth set threshold, the optimization mode selection module 5 is executed.
And the optimization mode selection module 5 is used for selecting a corresponding optimization mode according to the similarity.
A second optimization module 6, configured to optimize the initial optimized image using the selected optimization mode.
And a returning module 7, configured to take the optimized image as an initial optimized image, and return to the "similarity calculation module 3".
As a preferred embodiment, the similarity calculation module 3 of the present invention specifically includes:
and the first area dividing unit is used for dividing the target identification area on the initial optimization image to obtain a first alignment area.
And the second area dividing unit is used for dividing the target identification area on the standard image to obtain a second alignment area.
The format unification processing unit is used for carrying out format unification processing on the first alignment area and the second alignment area to obtain a third alignment area and a fourth alignment area; the third alignment region and the fourth alignment region both comprise mxn pixel points; wherein m and n are both positive integers greater than 5.
And the similarity calculation unit is used for calculating the similarity according to each pixel point in the third alignment area and each pixel point in the fourth alignment area.
As a preferred embodiment, the similarity calculation unit of the present invention specifically includes:
and the mean value calculating subunit is configured to calculate a mean value according to the pixel mean value corresponding to the fourth alignment area.
And the assignment subunit is configured to set, to 1, each pixel point that is greater than or equal to the mean value, and set, to 0, each pixel point that is smaller than the mean value, so as to obtain a first transformation matrix corresponding to the third alignment area and a second transformation matrix corresponding to the fourth alignment area.
And the similarity calculation subunit is used for calculating the similarity according to the first transformation matrix and the second transformation matrix.
As a preferred embodiment, the optimization mode selection module 5 of the present invention specifically includes:
the first judging unit is used for judging whether the similarity belongs to a first set threshold range or not; if the current time is within the first set threshold range, adopting a first mode as an optimization mode; the first mode is to perform illumination enhancement operation on the image.
The second judging unit is used for judging whether the similarity belongs to a second set threshold range or not; if the current value belongs to the range of the second set threshold, adopting a mode two as an optimization mode; the second mode is bilateral filtering and correction operation.
The third judging unit is used for judging whether the similarity belongs to a third set threshold range or not; if the current time is within the range of a third set threshold, adopting a mode three as an optimization mode; and the third mode is to perform exposure enhancement operation on the image.
A fourth judging unit, configured to judge whether the similarity falls within a fourth set threshold range; if the current time is within the range of the fourth set threshold, adopting a mode four as an optimization mode; and the fourth mode is to perform stretching operation on the saturation component.
The same portions as those in embodiment 1 are specifically referred to in embodiment 1, and are not described in detail herein.
In the description of the present invention, the description of the terms "one embodiment," "some embodiments," "specific embodiments," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present invention. In the present invention, the schematic representations of the terms used above do not necessarily refer to the same embodiment or example.
The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (10)
1. A method of image optimization, the method comprising:
acquiring an image to be optimized and a standard image;
inputting the image to be optimized into an image optimal model for optimization to obtain an initial optimized image;
calculating the similarity between the initial optimized image and the standard image;
judging whether the similarity belongs to a fifth set threshold range or not; if the image belongs to the range of a fifth set threshold, taking the initial optimized image as a final optimized image; if the similarity does not belong to the range of a fifth set threshold, selecting a corresponding optimization mode according to the similarity;
optimizing the initial optimized image by using the selected optimization mode;
and taking the optimized image as an initial optimized image, and returning to the step of calculating the similarity between the initial optimized image and the standard image.
2. The image optimization method according to claim 1, further comprising, before the step of "acquiring the image to be optimized and the standard image":
acquiring a sample image set obtained by acquiring target objects by a plurality of different types of image collectors and an annotation image set corresponding to the sample image;
selecting a training set from the sample image set and the labeled image set according to a first set proportion; selecting a test set from the rest of the sample image sets and the rest of the marked image sets according to a second set proportion;
training a neural network by using the training set to obtain an image initial optimization model;
verifying the initial image optimization model by using the test set to obtain accuracy;
judging whether the accuracy reaches a set value, and if the accuracy is greater than or equal to the set value, taking the image initial optimization model as an image optimal model; and if the accuracy is smaller than the set value, reselecting the training set and the test set.
3. The image optimization method according to claim 1, wherein the step of calculating the similarity between the initial optimized image and the standard image comprises:
dividing a target identification area on the initial optimization image to obtain a first alignment area;
dividing a target identification area on the standard image to obtain a second alignment area;
carrying out format unification processing on the first alignment area and the second alignment area to obtain a third alignment area and a fourth alignment area; the third alignment region and the fourth alignment region both comprise mxn pixel points; wherein m and n are both positive integers greater than 5;
and calculating the similarity according to each pixel point in the third alignment area and each pixel point in the fourth alignment area.
4. The image optimization method according to claim 3, wherein the calculating of the similarity according to each pixel point in the third alignment region and each pixel point in the fourth alignment region comprises:
calculating a mean value according to the pixel mean value corresponding to the fourth alignment area;
setting all the pixel points which are larger than or equal to the mean value as 1, and setting all the pixel points which are smaller than the mean value as 0, and obtaining a first transformation matrix corresponding to the third alignment area and a second transformation matrix corresponding to the fourth alignment area;
and calculating the similarity according to the first transformation matrix and the second transformation matrix.
5. The image optimization method according to claim 1, wherein the selecting a corresponding optimization mode according to the similarity specifically includes:
judging whether the similarity belongs to a first set threshold range or not; if the current time is within the first set threshold range, adopting a first mode as an optimization mode; the first mode is to perform illumination enhancement operation on an image;
judging whether the similarity belongs to a second set threshold range; if the current time is within the range of a second set threshold, adopting a mode two as an optimization mode; the second mode is bilateral filtering and correcting operation;
judging whether the similarity belongs to a third set threshold range or not; if the current time is within the range of a third set threshold, adopting a mode three as an optimization mode; the third mode is to carry out exposure enhancement operation on the image;
judging whether the similarity belongs to a fourth set threshold range or not; if the current time is within the range of the fourth set threshold, adopting a mode four as an optimization mode; and the fourth mode is to perform stretching operation on the saturation component.
6. An image optimization method according to claim 1, further comprising, after said optimizing the initial optimized image using the selected optimization mode:
judging whether the optimization times reach an optimization time threshold value or not; if the optimization times reach the threshold value of the optimization times, taking the initial optimization image as a final optimization image; and if the optimization times do not reach the threshold value of the optimization times, executing the step of taking the optimized image as the initial optimized image.
7. An image optimization system, characterized in that the system comprises:
the acquisition module is used for acquiring an image to be optimized and a standard image;
the first optimization module is used for inputting the image to be optimized into an image optimal model for optimization to obtain an initial optimized image;
the similarity calculation module is used for calculating the similarity between the initial optimized image and the standard image;
the judging module is used for judging whether the similarity belongs to a fifth set threshold range or not; if the image belongs to a fifth set threshold range, taking the initial optimized image as a final optimized image; if the current time does not fall into the range of the fifth set threshold value, executing an optimization mode selection module;
the optimization mode selection module is used for selecting a corresponding optimization mode according to the similarity;
the second optimization module is used for optimizing the initial optimization image by using the selected optimization mode;
and the returning module is used for taking the optimized image as an initial optimized image and returning to the similarity calculating module.
8. The image optimization system according to claim 7, wherein the similarity calculation module specifically includes:
a first area dividing unit, configured to perform target identification area division on the initial optimized image to obtain a first alignment area;
the second area dividing unit is used for dividing the target identification area on the standard image to obtain a second alignment area;
the format unification processing unit is used for carrying out format unification processing on the first alignment area and the second alignment area to obtain a third alignment area and a fourth alignment area; the third alignment region and the fourth alignment region both comprise mxn pixel points; wherein m and n are both positive integers greater than 5;
and the similarity calculation unit is used for calculating the similarity according to each pixel point in the third alignment area and each pixel point in the fourth alignment area.
9. The image optimization system according to claim 8, wherein the similarity calculation unit specifically includes:
the mean value calculating subunit is configured to calculate a mean value according to the pixel mean value corresponding to the fourth alignment area;
the assignment subunit is configured to set, to 1, each pixel point that is greater than or equal to the mean value, and set, to 0, each pixel point that is smaller than the mean value, so as to obtain a first transformation matrix corresponding to the third alignment area and a second transformation matrix corresponding to the fourth alignment area;
and the similarity calculation subunit is used for calculating the similarity according to the first transformation matrix and the second transformation matrix.
10. The image optimization system according to claim 7, wherein the optimization mode selection module specifically includes:
the first judging unit is used for judging whether the similarity belongs to a first set threshold range or not; if the current time is within the first set threshold range, adopting a first mode as an optimization mode; the first mode is to perform illumination enhancement operation on an image;
the second judging unit is used for judging whether the similarity belongs to a second set threshold range or not; if the current time is within the range of a second set threshold, adopting a mode two as an optimization mode; the second mode is bilateral filtering and correcting operation;
a third judging unit, configured to judge whether the similarity falls within a third set threshold range; if the current time is within the range of a third set threshold, adopting a mode three as an optimization mode; the third mode is to perform exposure enhancement operation on the image;
a fourth judging unit, configured to judge whether the similarity falls within a fourth set threshold range; if the current time is within the range of the fourth set threshold, adopting a mode four as an optimization mode; and the fourth mode is to perform stretching operation on the saturation component.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211417961.3A CN115760609B (en) | 2022-11-14 | 2022-11-14 | Image optimization method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211417961.3A CN115760609B (en) | 2022-11-14 | 2022-11-14 | Image optimization method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115760609A true CN115760609A (en) | 2023-03-07 |
| CN115760609B CN115760609B (en) | 2023-07-28 |
Family
ID=85370135
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211417961.3A Active CN115760609B (en) | 2022-11-14 | 2022-11-14 | Image optimization method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115760609B (en) |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107292830A (en) * | 2016-12-28 | 2017-10-24 | 首都师范大学 | Low-light (level) image enhaucament and evaluation method |
| CN108198144A (en) * | 2017-12-28 | 2018-06-22 | 广东欧珀移动通信有限公司 | Image processing method, device, storage medium and electronic equipment |
| CN109948653A (en) * | 2019-02-15 | 2019-06-28 | 广东智媒云图科技股份有限公司 | A kind of image similarity judgment method, electronic equipment and storage medium |
| CN110111286A (en) * | 2019-05-16 | 2019-08-09 | 北京印刷学院 | The determination method and apparatus of image optimization mode |
| CN112634170A (en) * | 2020-12-30 | 2021-04-09 | 平安科技（深圳）有限公司 | Blurred image correction method and device, computer equipment and storage medium |
| CN114219731A (en) * | 2021-12-15 | 2022-03-22 | 北京百度网讯科技有限公司 | Picture optimization method, device, equipment and storage medium |
| WO2022062828A1 (en) * | 2020-09-23 | 2022-03-31 | 深圳云天励飞技术股份有限公司 | Image model training method, image processing method, chip, device and medium |
| CN114723705A (en) * | 2022-03-31 | 2022-07-08 | 海门市恒创织带有限公司 | Cloth flaw detection method based on image processing |
| CN114757214A (en) * | 2022-05-12 | 2022-07-15 | 北京百度网讯科技有限公司 | Selection method and related device for sample corpora for optimizing translation model |
| CN115273114A (en) * | 2022-07-21 | 2022-11-01 | 厦门哲林软件科技有限公司 | Document image optimization method and medium |
-
2022
- 2022-11-14 CN CN202211417961.3A patent/CN115760609B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107292830A (en) * | 2016-12-28 | 2017-10-24 | 首都师范大学 | Low-light (level) image enhaucament and evaluation method |
| CN108198144A (en) * | 2017-12-28 | 2018-06-22 | 广东欧珀移动通信有限公司 | Image processing method, device, storage medium and electronic equipment |
| CN109948653A (en) * | 2019-02-15 | 2019-06-28 | 广东智媒云图科技股份有限公司 | A kind of image similarity judgment method, electronic equipment and storage medium |
| CN110111286A (en) * | 2019-05-16 | 2019-08-09 | 北京印刷学院 | The determination method and apparatus of image optimization mode |
| WO2022062828A1 (en) * | 2020-09-23 | 2022-03-31 | 深圳云天励飞技术股份有限公司 | Image model training method, image processing method, chip, device and medium |
| CN112634170A (en) * | 2020-12-30 | 2021-04-09 | 平安科技（深圳）有限公司 | Blurred image correction method and device, computer equipment and storage medium |
| CN114219731A (en) * | 2021-12-15 | 2022-03-22 | 北京百度网讯科技有限公司 | Picture optimization method, device, equipment and storage medium |
| CN114723705A (en) * | 2022-03-31 | 2022-07-08 | 海门市恒创织带有限公司 | Cloth flaw detection method based on image processing |
| CN114757214A (en) * | 2022-05-12 | 2022-07-15 | 北京百度网讯科技有限公司 | Selection method and related device for sample corpora for optimizing translation model |
| CN115273114A (en) * | 2022-07-21 | 2022-11-01 | 厦门哲林软件科技有限公司 | Document image optimization method and medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115760609B (en) | 2023-07-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107578390B (en) | Method and device for correcting image white balance by using neural network | |
| CN108875821A (en) | The training method and device of disaggregated model, mobile terminal, readable storage medium storing program for executing | |
| CN107516319B (en) | High-precision simple interactive matting method, storage device and terminal | |
| CN111310756B (en) | Damaged corn particle detection and classification method based on deep learning | |
| CN110400278B (en) | Full-automatic correction method, device and equipment for image color and geometric distortion | |
| CN111027382B (en) | Attention mechanism-based lightweight face detection method and model | |
| CN111986785B (en) | Medical image labeling method, device, equipment and storage medium | |
| CN109903282B (en) | Cell counting method, system, device and storage medium | |
| CN114972085B (en) | Fine granularity noise estimation method and system based on contrast learning | |
| CN112884782B (en) | Biological object segmentation method, apparatus, computer device, and storage medium | |
| CN110991256A (en) | System and method for carrying out age estimation and/or gender identification based on face features | |
| CN108734667B (en) | Image processing method and system | |
| CN116757988A (en) | Infrared and visible light image fusion method based on semantic enrichment and segmentation tasks | |
| CN117496019B (en) | Image animation processing method and system for driving static image | |
| CN114120359A (en) | Method for measuring body size of group-fed pigs based on stacked hourglass network | |
| CN117541574A (en) | Tongue diagnosis detection method based on AI semantic segmentation and image recognition | |
| CN113283466A (en) | Instrument reading identification method and device and readable storage medium | |
| CN115830514B (en) | Whole river reach surface flow velocity calculation method and system suitable for curved river channel | |
| CN115760609B (en) | Image optimization method and system | |
| CN113112531B (en) | Image matching method and device | |
| CN111260625B (en) | Automatic extraction method for offset printing large image detection area | |
| CN114358131A (en) | Digital photo frame intelligent photo optimization processing system | |
| CN111753835B (en) | Cell tracking method based on local graph matching and convolutional neural network | |
| CN114758002A (en) | Photovoltaic string position determining method and system based on aerial image | |
| JP2004199200A (en) | Pattern recognition device, imaging apparatus, information processing system, pattern recognition method, recording medium and program |
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 | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20230705 Address after: No. 62, Xialu Lane, Nanda Road, Jiaocheng District, Ningde City, Fujian Province, 352000 Applicant after: Lin Yaxia Address before: 541000 Information Industry Park, Chaoyang Township, Qixing District, Guilin, Guangxi Zhuang Autonomous Region Applicant before: Wang Yuxin |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20231121 Address after: Room 401 and Room 402B, No. 1280 Yinxing Road, Yangpu District, Shanghai, 200082 Patentee after: Aide Linker (Shanghai) Digital Technology Co.,Ltd. Address before: No. 62, Xialu Lane, Nanda Road, Jiaocheng District, Ningde City, Fujian Province, 352000 Patentee before: Lin Yaxia |