CN109035161A - A kind of reconstructing method and device of image data - Google Patents
A kind of reconstructing method and device of image data Download PDFInfo
- Publication number
- CN109035161A CN109035161A CN201810723759.0A CN201810723759A CN109035161A CN 109035161 A CN109035161 A CN 109035161A CN 201810723759 A CN201810723759 A CN 201810723759A CN 109035161 A CN109035161 A CN 109035161A
- Authority
- CN
- China
- Prior art keywords
- parameter
- data
- template
- value
- image data
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 29
- 230000000694 effects Effects 0.000 abstract description 6
- 241000208340 Araliaceae Species 0.000 description 4
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 4
- 235000003140 Panax quinquefolius Nutrition 0.000 description 4
- 235000008434 ginseng Nutrition 0.000 description 4
- 238000010586 diagram Methods 0.000 description 2
- 238000009434 installation Methods 0.000 description 2
- 241001269238 Data Species 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/50—Image enhancement or restoration using two or more images, e.g. averaging or subtraction
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
Abstract
The present invention relates to a kind of reconstructing methods of image data, include the following steps: the template of setting processing image data；Select its data as original data in any one frame image；Select the image data on the corresponding position in the next frame image in currently reception or conventional images as movement images data；Compare each of two data parameter one by one, it is such as identical, then select the value for the reconstruction value of the parameter；If not identical, then operation is carried out to two data, and one of reconstruction value of the value as the parameter is selected according to the result of the operation；By the parameter reconstruction value of acquirement according to the format combination of setting after traversing all parameters in template, the image data reconstructed.The invention further relates to a kind of devices for realizing the above method.The reconstructing method and device for implementing a kind of image data of the invention have the advantages that it not and will receive that the limitation of image frames numbers, effect are preferable.
Description
Technical field
The present invention relates to image real time transfers, more specifically to the reconstructing method and device of a kind of image data.
Background technique
When video field handles image, the certain parts that some images in most cases can all occur are unclear
The case where Chu.In general, in the prior art, these unclear image-erasings can be left at clearly image
Reason, because the quantity of image in general is enough.But in some cases, such as in some special occasions or spy
Different stage, such processing method will go wrong, at this time may image quantity or the frame number of image and few, if
Unclear picture frame is excluded except process range or deleted, the number of image frames for participating in handling will be made inadequate, thus
Prevent treated image is from accurately expressing true environment.Therefore, existing image real time transfer mode is in some fields
The using effect in conjunction or stage is not that very well, cannot restore practical situation.
Summary of the invention
The technical problem to be solved in the present invention is that the above-mentioned use in some cases for the prior art is schemed
As the bad defect of the limitation of frame number, treatment effect, a kind of limitation, effect preferable one that not will receive image frames numbers is provided
The reconstructing method and device of kind image data.
The technical solution adopted by the present invention to solve the technical problems is: constructing a kind of reconstructing method of image data, wraps
Include following steps:
A) size, parameter and its format of the template of setting processing image data；
B) any position in any one frame image selects its data as initial pictures number according to the template size
According to；
C it) is selected according to the template size on the corresponding position in the next frame image in current reception or conventional images
Image data as movement images data；
D) each of the original data and the movement images data parameter one by one, it is such as identical, then
Select the value for the reconstruction value of the parameter；If not identical, then operation is carried out to two data, and select according to the result of the operation
Select one of reconstruction value of the value as the parameter；
E) judge whether all parameters in the template traverse, if so, by the parameter reconstruction value of acquirement according to setting
Format combination, the image data reconstructed；Otherwise, return step D) and carry out the comparison of next parameter.
Further, further include following steps when size of the template size less than a frame image data:
F) selected in any one frame image different positions according to the size of the template select its data as
The original data, return step C), and repeating said steps C) arrive E), until traversing any one frame image.
Further, further include following steps when the template size is equal to the size of a frame image data:
G) using the image data of obtained reconstruct as the original data, return step C), and repeat the step
Rapid C) arrive E).
Further, the step A) in, the size parameter of the template includes the pixel quantity for including in the template,
The parameter and format include the parameter value that the pixel characteristic is determined in the position and each pixel of each pixel；Wherein, mould
Include multiple pixels in plate, a pixel includes multiple descriptions or limits the parameter of its feature, in template pixel quantity be less than or
Equal to pixel quantity in a frame image data.
Further, the step D) in, when participating in the parameter value difference compared for two, whether judge its diversity factor
Greater than given threshold, if so, selecting a parameter value as reconstruct in the parameter for participating in comparing, otherwise, selection is wherein first
Parameter value in beginning image is as reconstruction parameter value.
Further, the diversity factor includes the ratio of two parameter values；The threshold value is the ratio value of a setting.
Further, the diversity factor and the threshold value are respectively less than 1；Described two parameter values compare its absolute value first
Then size is divided by according to the position of setting, obtain the difference angle value less than 1.
The invention further relates to a kind of devices for realizing the above method, comprising:
Template-setup module: for size, parameter and its format of the template of processing image data to be arranged；
Original data selecting module: it is selected for any position in any one frame image according to the template size
Its data is selected as original data；
Movement images data selecting module: under being selected in current reception or conventional images according to the template size
The image data on corresponding position in one frame image is as movement images data；
Comparison module: each of the original data and the movement images data parameter one by one, such as
It is identical, then select the value for the reconstruction value of the parameter；If not identical, then operation is carried out to two data, and according to the operation
Result select the one of reconstruction value of value as the parameter；
Judgment module: whether all parameters for judging in the template traverse, if so, by the parameter reconstruction value of acquirement
According to the format combination of setting, the image data that is reconstructed；Otherwise, comparison module is returned, and carries out the ratio of next parameter
Compared with.
Further, further includes:
Frame processing module: for when size of the template size less than a frame image data, in any one frame
It selects different positions to select its data as the original data according to the size of the template in image, and carries out weight
Structure processing, until traversing any one frame image；
Iteration module: the image data of the frame reconstruct for that will obtain carries out weight as the original data
Structure processing.
Further, the size parameter of the template includes the pixel quantity for including, the parameter and lattice in the template
Formula includes the parameter value that the pixel characteristic is determined in the position and each pixel of each pixel；It wherein, include multiple in template
Pixel, a pixel include multiple descriptions or the parameter for limiting its feature, and pixel quantity is less than or equal to a frame image in template
Pixel quantity in data.
The reconstructing method and device for implementing a kind of image data of the invention have the advantages that due to by image
Data according to setting template size obtain, and using successive two frame image same position data parameters into
Row compare, if the value of a parameter be in the data between successive two frame it is identical, this parameter value will not be changed；Such as
Fruit is different, then one is found in the parameter value for being belonging respectively to two different frames more close to virtual value or true value by operation
Parameter value, as reconstruct when the parameter value so that reconstruct after image data more close to true environment generate number
According to.In this way, comparing by one or many templates, the reconstruct data of a frame image can be obtained；The data of multiple image are logical
It crosses aforesaid way to be iterated, it will be able to obtain true ambient image data.Therefore, the limit of image frames numbers is not will receive
System, effect are preferable.
Detailed description of the invention
Fig. 1 is a kind of reconstructing method of image data of the present invention and the structural schematic diagram of Installation practice；
Fig. 2 is the apparatus structure schematic diagram in the embodiment.
Specific embodiment
Below in conjunction with attached drawing, embodiments of the present invention is further illustrated.
As shown in Figure 1, in a kind of reconstructing method of image data of the invention and Installation practice, the image data
Reconstructing method includes the following steps:
Template parameter is arranged in step S11: in the present embodiment, already existing or just in received image with processing a batch
Data instance illustrates the data reconstruction method of image procossing.Wherein, these image datas be all in the form of picture frame exist or
Storage.In general, image data refers to the data that image can be finally shown as by image display software, these data with
One frame is unit presence, and is arranged in a certain order, image data one data set of formation of a frame, in the data set not
The only number including multiple pixels, its position etc. further include definition or the parameter for limiting the pixel, for example, the display
The color of display, brightness etc..In this step, setting handles size, parameter and its format of the template of image data, is exactly
The size and format for defining the data volume of single treatment in this method, so that the physical significance of each variable therein or parameter is bright
Really.Specifically, it is of course possible to select the size of a frame as the size of template, processing speed is very fast in this case, energy
It enough receives a frame image data just to be handled, but in this case, all for the speed of processor, the quantity of caching
There is higher requirement, is not that can use under any circumstance；And in yet some other cases, it also can choose template size
It is smaller, only a part of a frame image data, for example, 1/8 or 1/16 etc., although its processing speed is slower, need through excessive
Secondary handle just is able to achieve the reconstruct of a frame image data, but requirement of this processing mode for hardware aspect is lower, can be with
The reduction on hardware cost is brought, insensitive to the processing time but in the case where to cost sensitivity, can obtain preferable
It uses.In short, in the present embodiment, the size parameter of the template includes the pixel quantity for including, the parameter in the template
It include the parameter value that the pixel characteristic is determined in the position and each pixel of each pixel with format；Wherein, include in template
Multiple pixels, a pixel include multiple descriptions or the parameter for limiting its feature, and pixel quantity is less than or equal to a frame in template
Pixel quantity in image data.It is noted that the processing when handling the data in template, to pixel therein
Ear algorithm etc. can be cut using recurrence under in some cases, to guarantee the covering for all pixels in module, meanwhile, in this way
Algorithm can also make the shape of template that there is more shapes, including arbitrary polygon, to guarantee template for frame
The covering of data.For generally, in the present embodiment, frame data is decomposed into pixel and its indicate the ginseng of the pixel characteristic
Number, and the correspondence parameter value of different frame is compared, and then achieve the purpose that correct frame data, it can make at entire data
Manage relatively simple, the processing time is shorter, and convergence is very fast, can reappear ambient image number as far as possible in the limited situation of initial data
According to.
Step S12 selects primary data: in this step, any position in any one frame image is according to the template
Size selects its data as original data；It in the present embodiment, essentially, is by the data of two picture frames
(for example, data of two adjacent picture frames) compare one by one, unintelligible for picture frame unsharp for part
Part be different certainly with the parameter of corresponding position of normal frame, will be unintelligible by the processing of method in the present embodiment
The unintelligible part of frame is substituted using the relevant parameter of normal frame or more normal frame, finally obtains all ginsengs in reconstruct data
Number is all normal, and then obtains a normal or clearly image frame data.When the size of template and a frame image data
When identical, the original data in this step is actually the frame data of selection；When template size is less than frame size, on
Stating optional position at least should make template itself without departing from frame data, and in other words, above-mentioned selection should be ensured that
Template is fully fallen among frame data.Preferably selection mode is such that in after integer time selection, can just cover one
Frame data.
Data are compared in step S13 selection, and compare wherein each parameter one by one: selecting according to the template size currently received
Or the image data on the corresponding position in the next frame image in conventional images is as movement images data, and selects wherein one
A parameter starts to compare.In the present embodiment, above-mentioned steps S12 has selected one in two data sets being compared, and
This step selects another in the two datasets being compared.Wherein, exist in this comparison data selected in step
Position in image frame data where it is corresponding or consistent with position of the above-mentioned primary data where it in picture frame.
For example, it is assumed that there is 16 pixels in an image frame data, each pixel is defined or is limited by 8 parameters, template it is big
Small is 4 pixels, has selected preceding 4 pixels in a frame as primary data in step s 12, has shared 32 parameters；At this
In step, equally select preceding 4 pixels in next frame data as data are compared, which equally has 32 ginsengs
Number, and each parameter is corresponding with a parameter in primary data.And when relatively, it is by above-mentioned primary data and to compare number
Parameter in compares one by one, for example, by first parameter in primary data and compare first parameter in data and compare,
By the second parameter in primary data and compare the second parameter in data and compare, until above-mentioned 32 parameters are respectively compared
It completes, then completes this template data and compare.In general, above-mentioned comparison be according to parameter in data set put in order into
Capable.
Step S14 parameter value is identical no: in this step, judging that participating in comparing, two respectively from above-mentioned initial number
It is whether identical according to the parameter with the same position for comparing data, it is such as identical, execute step S15；If not identical, execution step S16.
Step S15 selects the value for the reconstruction value of the parameter: in this step, select the value for the reconstruction value of the parameter,
After having executed this step, step S17 is executed.
Step S16 carries out operation to two parameter values comparing of participation, and according to operation result select it is one of as
The reconstruction value of the parameter: in this step, due to the parameter value of two comparisons and different, therefore carrying out operations to two data,
And one of reconstruction value of the value as the parameter is selected according to the result of the operation, then execute step S17.As one
For preferred example, in the present embodiment, its diversity factor can be judged again when participating in the parameter value difference compared for two
Whether given threshold is greater than, if so, selecting parameter of any one value as reconstruct between two parameter values for participating in comparing
Value, otherwise, selects the parameter value in wherein initial pictures as reconstruction parameter value.It is noted that in the present embodiment, institute
State the ratio that diversity factor includes two parameter values；The threshold value is the ratio value of a setting.The diversity factor and the threshold value
Respectively less than 1；Described two parameter values compare its order of magnitude first, are then divided by according to the position of setting, obtain less than 1
Difference angle value.That is, in this step, when the parameter value that two participations are compared is not identical, the two will be compared first
Then the size of parameter absolute value can will be worth lesser one as molecule, biggish one is used as denominator, obtains less than 1
Diversity factor, 1 threshold value comparison is similarly less than be previously set one using the obtained diversity factor, for example, 0.5, judgement
Whether above-mentioned diversity factor is less than the threshold value, if so, indicating the difference of two parameters and little, the parameter value in selection initial pictures
As reconstruction parameter value；If not, select one in two parameter values as reconstruction parameter value.For in two values or element
Middle selection is more effective to be selected as one, there are many existing algorithm or method it is optional, for example, can be using certain quickly
Sort method selection can also use such as Markov chain Monte-Carlo algorithm, the parameter that can also compare according to last time
The correspondence parameter value of value or nearest pixel point select etc..
Each parameter that step S17 is traversed in template is no, if so, executing step S18, carries out whether picture frame handles completion
Judgement；Otherwise, return step S13 and the comparison of next parameter is carried out.
It is no that step S18 has handled a frame data, such as processed, thens follow the steps S18；Otherwise, in any one frame
Different positions is selected to select its data as the original data, return step according to the size of the template in image
S12, and repeat the above steps.It is noted that it is this without the case where having handled a frame data, only appear in institute
When stating size of the template size less than a frame image data.When the above-mentioned arbitrary image frame for being selected as primary data acquirement of return
It when data select primary data again, will be selected according to above-mentioned record, i.e., just cover a frame image after integer time selection
Data are whole, it is, of course, also possible to the size of template is adjusted to cooperate this selection, i.e. in some cases, it can be at the beginning
This effect is just considered when setting template size, adjusts template size intentionally, in the hope of just covering after above-mentioned integer time selection
The case where one frame image data.
Step S19 obtains reconstructed image data using reconstruction value: in this step, by the parameter reconstruction value of acquirement according to setting
Fixed format combination, the image data reconstructed.
Above-mentioned steps show the concrete operations for obtaining frame reconstruct data.But in actual use, may not be
The unclear problem of image section can be eliminated by a frame reconstructed image data.In this case, it often requires repeatedly into
The above-mentioned reconstruct of row could be solved the problems, such as preferably.At this point, further including following steps in the present embodiment: by the image of obtained reconstruct
Data repeat the above steps as the original data, return step S12, until the picture frame of reconstruct is all clear.
It in some sense, is equivalent to and (in a period of time) all images frame is superimposed, and to all parameters therein
Meet it is above-mentioned as defined in iteration, a whole parameters are all clear, effective image frame data until obtaining, that is, obtain
One can completely react the picture frame of currently practical environment.
A kind of device for realizing the above method is further related in the present embodiment, refers to Fig. 2, which includes: that template is set
Set module 1, original data selecting module 2, movement images data selecting module 3, comparison module 4, judgment module 5 and frame
Processing module 6.Wherein, template-setup module 1 is used to be arranged size, parameter and its format of the template of processing image data；Just
Beginning image data selecting module 2 selects its data to make for any position in any one frame image according to the template size
For original data；Movement images data selecting module 3 is used to select current reception or existing figure according to the template size
The image data on the corresponding position in next frame image as in is as movement images data；Comparison module 4 for comparing one by one
Each of the original data and the movement images data parameter, it is such as identical, then select the value for the parameter
Reconstruction value；If not identical, then operations are carried out to two data, and according to the result of the operation select one of them value as
The reconstruction value of the parameter；Judgment module 5 is for judging whether all parameters in the template traverse, if so, by the ginseng of acquirement
Format combination of the number reconstruction value according to setting, the image data reconstructed；Otherwise, comparison module is returned, and is carried out next
The comparison of parameter；
Frame processing module 6 is used for when size of the template size less than a frame image data, in any one frame
It selects different positions to select its data as the original data according to the size of the template in image, and carries out weight
Structure processing, until traversing any one frame image；
In addition, the device further includes iteration module 7, the image data for the frame reconstruct which is used to obtain
As the original data, and carry out reconstruction processing.Wherein, to include in the template include the size parameter of the template
Pixel quantity, the parameter and format include the parameter that the pixel characteristic is determined in the position and each pixel of each pixel
Value；It wherein, include multiple pixels in template, a pixel includes multiple descriptions or the parameter for limiting its feature, pixel in template
Quantity is less than or equal to pixel quantity in a frame image data.
The embodiments described above only express several embodiments of the present invention, and the description thereof is more specific and detailed, but simultaneously
Limitations on the scope of the patent of the present invention therefore cannot be interpreted as.It should be pointed out that for those of ordinary skill in the art
For, without departing from the inventive concept of the premise, various modifications and improvements can be made, these belong to guarantor of the invention
Protect range.Therefore, the scope of protection of the patent of the invention shall be subject to the appended claims.
Claims (10)
1. a kind of reconstructing method of image data, which comprises the steps of:
A) size, parameter and its format of the template of setting processing image data；
B) any position in any one frame image selects its data as original data according to the template size；
C the figure on the corresponding position in the next frame image in current reception or conventional images) is selected according to the template size
As data are as movement images data；
D) each of the original data and the movement images data parameter one by one, it is such as identical, then it selects
The value is the reconstruction value of the parameter；If not identical, then operation is carried out to two data, and it is selected according to the result of the operation
In a reconstruction value of the value as the parameter；
E) judge whether all parameters in the template traverse, if so, by the parameter reconstruction value of acquirement according to the format of setting
Combination, the image data reconstructed；Otherwise, return step D) and carry out the comparison of next parameter.
2. the reconstructing method of image data according to claim 1, which is characterized in that when the template size is less than a frame
Further include following steps when the size of image data:
F) different positions is selected to select its data as described according to the size of the template in any one frame image
Original data, return step C), and repeating said steps C) arrive E), until traversing any one frame image.
3. the reconstructing method of image data according to claim 1, which is characterized in that further include following steps:
G) using the image data of obtained reconstruct as the original data, return step C), and repeating said steps C)
To E).
4. the reconstructing method of image data according to claim 1 to 3, which is characterized in that the step A)
In, the size parameter of the template includes the pixel quantity for including in the template, and the parameter and format include each pixel
The parameter value of the pixel characteristic is determined in position and each pixel；It wherein, include multiple pixels in template, a pixel includes
Multiple descriptions limit the parameters of its feature, and pixel quantity is less than or equal to pixel quantity in a frame image data in template.
5. the reconstructing method of image data according to claim 4, which is characterized in that the step D) in, join at two
When parameter value difference compared with, judge whether its diversity factor is greater than given threshold, if so, selection participates in the parameter compared
Otherwise one parameter value as reconstruct selects the parameter value in wherein initial pictures as reconstruction parameter value.
6. the reconstructing method of image data according to claim 5, which is characterized in that the diversity factor includes two parameters
The ratio of value；The threshold value is the ratio value of a setting.
7. the reconstructing method of image data according to claim 6, which is characterized in that the diversity factor and the threshold value are equal
Less than 1；Described two parameter values compare its order of magnitude first, are then divided by according to the position of setting, obtain the difference less than 1
Different angle value.
8. a kind of device for the reconstructing method for realizing image data as described in claim 1 characterized by comprising
Template-setup module: for size, parameter and its format of the template of processing image data to be arranged；
Original data selecting module: it is selected according to the template size for any position in any one frame image
Data are as original data；
Movement images data selecting module: for selecting the next frame in current reception or conventional images according to the template size
The image data on corresponding position in image is as movement images data；
Comparison module: each of the original data and the movement images data parameter one by one, it is such as identical,
Then select the value for the reconstruction value of the parameter；If not identical, then operation is carried out to two data, and according to the result of the operation
Select one of reconstruction value of the value as the parameter；
Judgment module: whether all parameters for judging in the template traverse, if so, by the parameter reconstruction value of acquirement according to
The format combination of setting, the image data reconstructed；Otherwise, comparison module is returned, and carries out the comparison of next parameter.
9. device according to claim 8, which is characterized in that further include:
Frame processing module: for when size of the template size less than a frame image data, in any one frame image
It is middle that different positions is selected to select its data as the original data according to the size of the template, and place is reconstructed
Reason, until traversing any one frame image；
Iteration module: place is reconstructed as the original data in the image data of the frame reconstruct for that will obtain
Reason.
10. device according to claim 8, which is characterized in that the size parameter of the template, which includes in the template, includes
Pixel quantity, the parameter and format include the parameter that the pixel characteristic is determined in the position and each pixel of each pixel
Value；It wherein, include multiple pixels in template, a pixel includes multiple descriptions or the parameter for limiting its feature, pixel in template
Quantity is less than or equal to pixel quantity in a frame image data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810723759.0A CN109035161A (en) | 2018-07-04 | 2018-07-04 | A kind of reconstructing method and device of image data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810723759.0A CN109035161A (en) | 2018-07-04 | 2018-07-04 | A kind of reconstructing method and device of image data |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109035161A true CN109035161A (en) | 2018-12-18 |
Family
ID=65521719
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810723759.0A Pending CN109035161A (en) | 2018-07-04 | 2018-07-04 | A kind of reconstructing method and device of image data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109035161A (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103795920A (en) * | 2014-01-21 | 2014-05-14 | 宇龙计算机通信科技(深圳)有限公司 | Photo processing method and device |
| US20150271522A1 (en) * | 2012-10-12 | 2015-09-24 | National Institute Of Information And Communications Technology | Method, program and apparatus for reducing data size of a plurality of images containing mutually similar information, and data structure representing a plurality of images containing mutually similar information |
| CN108063920A (en) * | 2017-12-26 | 2018-05-22 | 深圳开立生物医疗科技股份有限公司 | A kind of freeze frame method, apparatus, equipment and computer readable storage medium |
-
2018
- 2018-07-04 CN CN201810723759.0A patent/CN109035161A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150271522A1 (en) * | 2012-10-12 | 2015-09-24 | National Institute Of Information And Communications Technology | Method, program and apparatus for reducing data size of a plurality of images containing mutually similar information, and data structure representing a plurality of images containing mutually similar information |
| CN103795920A (en) * | 2014-01-21 | 2014-05-14 | 宇龙计算机通信科技(深圳)有限公司 | Photo processing method and device |
| CN108063920A (en) * | 2017-12-26 | 2018-05-22 | 深圳开立生物医疗科技股份有限公司 | A kind of freeze frame method, apparatus, equipment and computer readable storage medium |
Non-Patent Citations (2)
| Title |
|---|
| 苗启广著: "《多传感器图像融合技术及应用》", 30 April 2014, 西安电子科技大学出版社 * |
| 蔡成涛等著: "《海洋环境下的计算机视觉技术》", 31 October 2015, 国防工业出版社 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Baldassarre et al. | Deep koalarization: Image colorization using cnns and inception-resnet-v2 | |
| WO2023092813A1 (en) | Swin-transformer image denoising method and system based on channel attention | |
| CN111798400B (en) | Non-reference low-illumination image enhancement method and system based on generation countermeasure network | |
| CN108898560B (en) | Core CT image super-resolution reconstruction method based on three-dimensional convolutional neural network | |
| CN114092330B (en) | Light-weight multi-scale infrared image super-resolution reconstruction method | |
| US11222211B2 (en) | Method and apparatus for segmenting video object, electronic device, and storage medium | |
| CN107464217B (en) | Image processing method and device | |
| CN112801901A (en) | Image deblurring algorithm based on block multi-scale convolution neural network | |
| CN109934826A (en) | A kind of characteristics of image dividing method based on figure convolutional network | |
| CN109543548A (en) | A kind of face identification method, device and storage medium | |
| CN110148088B (en) | Image processing method, image rain removing method, device, terminal and medium | |
| CN112771578B (en) | Image generation using subdivision scaling and depth scaling | |
| CN109472193A (en) | Method for detecting human face and device | |
| CN110827297A (en) | Insulator segmentation method for generating countermeasure network based on improved conditions | |
| CN104253929B (en) | Vedio noise reduction method and its system | |
| CN113689517B (en) | Image texture synthesis method and system for multi-scale channel attention network | |
| CN108734677B (en) | Blind deblurring method and system based on deep learning | |
| CN110062282A (en) | A kind of super-resolution video method for reconstructing, device and electronic equipment | |
| Vecchio et al. | Surfacenet: Adversarial svbrdf estimation from a single image | |
| CN103489174A (en) | Human face super-resolution method based on residual keeping | |
| CN111626960A (en) | Image defogging method, terminal and computer storage medium | |
| Chen et al. | Automatic colorization of images from Chinese black and white films based on CNN | |
| Zhao et al. | Detecting deepfake video by learning two-level features with two-stream convolutional neural network | |
| CN114299573A (en) | Video processing method and device, electronic equipment and storage medium | |
| Kumar et al. | A Robust Approach for Image Super-Resolution using Modified Very Deep Convolution Networks |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181218 |