CN109191371A - A method of it judging automatically scenery type and carries out image filters processing - Google Patents
A method of it judging automatically scenery type and carries out image filters processing Download PDFInfo
- Publication number
- CN109191371A CN109191371A CN201810927477.2A CN201810927477A CN109191371A CN 109191371 A CN109191371 A CN 109191371A CN 201810927477 A CN201810927477 A CN 201810927477A CN 109191371 A CN109191371 A CN 109191371A
- Authority
- CN
- China
- Prior art keywords
- classification
- scenery
- neural network
- out image
- real
- 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.)
- Withdrawn
Links
- 238000012545 processing Methods 0.000 title claims abstract description 15
- 238000000034 method Methods 0.000 title claims abstract description 13
- 238000013528 artificial neural network Methods 0.000 claims abstract description 13
- 238000012549 training Methods 0.000 claims abstract description 4
- 238000003384 imaging method Methods 0.000 claims description 7
- 238000013135 deep learning Methods 0.000 claims description 3
- 210000004218 nerve net Anatomy 0.000 claims description 2
- 238000012956 testing procedure Methods 0.000 claims 1
- 230000000694 effects Effects 0.000 abstract description 6
- 238000009877 rendering Methods 0.000 abstract description 5
- 230000004927 fusion Effects 0.000 abstract description 2
- 230000001537 neural effect Effects 0.000 abstract description 2
- 238000003062 neural network model Methods 0.000 abstract 1
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000003672 processing method Methods 0.000 description 2
- 238000013145 classification model Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformations in the plane of the image
- G06T3/04—Context-preserving transformations, e.g. by using an importance map
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
The present invention provides a kind of methods for judging automatically scenery type and carrying out image filters processing, different filters is loaded according to different scenery based on neural fusion, one neural network model of training, this model can identify eight different scenery in real-time/later period: such as landscape, interior, portrait (makeup), personage (whole body collective), skin makeup, food, one is selected in the preset filter of each scenery based on the scenery that neural network feedback returns to be rendered, and more targetedly scenery filter rendering effect can be reached.
Description
Technical field
The present invention relates to digital picture filter Rendering, specifically a kind of scenery type that judges automatically carries out image filters
The method of processing.
Background technique
In the overall technology of digital picture filter rendering, for scenery selection, targetedly filter seems most important.
The effect of beautification food is not only not achieved if the mill skin filter of portrait is used for food in the one many classes of scenery point to shoot,
The original color filtration of food can also be fallen；Can only be for a kind of scenery if two to preset filter, it cannot be according to different scenery
It is targetedly rendered, it is difficult to meet the needs of users.
Mostly have many filters in camera applications in the market, such as portrait makeup, portrait U.S. face, food, wind
There are many filter of scape, and the filter that some cameras also have by it is very popular.
But temporarily there is not a application that can beautify personage up to now and beautify food, can be cut for special scenes
The application of filter is changed, is all relatively simple camera applications, such as there are a lot of filters for portrait U.S. face, but after user
When setting camera progress landscape shooting, it is landscape that system, which cannot recognize this scenery, is still rendered using the filter of U.S. face,
The effect to glorify the landscape is not achieved.Therefore for most users in order to meet the needs of oneself more scene shot, mobile phone the inside is average
Be equipped with 2-3 money camera applications, and using when need multiple applications to switch over, it is cumbersome.
Summary of the invention
The present invention is intended to provide it is a kind of judge automatically scenery type carry out image filters processing method so that shooting when pair
Different scenery can more have for filter rendering is carried out capablely, and landscaping effect is more preferably.
A method of it judging automatically scenery type and carries out image filters processing, comprising the following steps:
1) Nature Photography or real-time imaging, are divided into several classifications；
2), above-mentioned classification is trained using the neural network in deep learning；
3), when real-time imaging or picture input, classification judgement is carried out by model, wherein the output of each classification
Probability is that a kind of classification of maximum probability is chosen between 0~1.0 is that real-time imaging or picture are classified；
4) the default filter for, loading the classification automatically according to classification carries out filter processing to image or video frame；
5), image or video frame of the output after 4) step process.
The present invention can effectively be classified the image of input or image using neural-network classification method, in turn
According to different article loads, more suitably processing filter, main advantage have:
1, the article of nature is divided into eight major class, each type all has between versatility and Lie Bie with biggish
Difference is conducive to the classification accuracy for improving neural network.
2, the landscaping effect that filter is conducive to optimize each article image is loaded automatically, such as the processing method and wind of food
The processing scheme of scape is usually different, by identifying type, loads different processing filters, can make the category
Image more aestheticism, true to nature and natural.
3, the present invention is simple easily holds, and is widely used, not only can be applied in the software of various image procossings, can also be
At present it is applied in digital filming product, and operational efficiency is very high, practicability is also very desirable.
Detailed description of the invention
Attached drawing is used to provide further understanding of the present invention, and constitutes part of specification, with reality of the invention
It applies example to be used to explain the present invention together, not be construed as limiting the invention.In the accompanying drawings:
Fig. 1 is flow diagram of the embodiment of the present invention.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall within the protection scope of the present invention.
As shown in Figure 1, a kind of technology for judging automatically scenery type and carrying out filter processing, detailed steps are as follows:
1) Nature Photography or real-time image are divided into following eight major class by the present invention:
2) above-mentioned classification is trained using the neural network in deep learning.It is main comprising steps of
A) according to eight above-mentioned big classifications, the other picture training sample of 8 major class is arranged respectively.
B) neural network is constructed, the sample in a) is input in neural network using the scheme of neural network classification
It is trained.
C) model that b) step obtains is tested, until model has enough classification capacities.
3) in software using neural network classification model obtained in 2) step.When real-time imaging or picture input
When, classification judgement is carried out by model.Wherein the output probability of each classification is to choose the one of maximum probability between 0~1.0
Kind, which is classified, determines that the image is to belong to any classification.
4) taxonomy of goods of the image or video frame will be obtained after 3) step, which is loaded according to classification automatically
Default filter filter processing is carried out to image or video frame.
5) image or video frame of the output after 4) step process.
The present invention predominantly loads different filters, one nerve net of training according to different scenery based on neural fusion
Network model, this model can identify eight different scenery in real-time/later period: such as landscape, interior, portrait (makeup), personage's (whole body collection
Body), skin makeup, food, the scenery returned based on neural network feedback selected one in the preset filter of each scenery and rendered,
More targetedly scenery filter rendering effect can be reached.
It although an embodiment of the present invention has been shown and described, for the ordinary skill in the art, can be with
A variety of variations, modification, replacement can be carried out to these embodiments without departing from the principles and spirit of the present invention by understanding
And modification, the scope of the present invention is defined by the appended.
Claims (3)
1. a kind of method for judging automatically scenery type and carrying out image filters processing, it is characterised in that the following steps are included:
1) Nature Photography or real-time imaging, are divided into several classifications；
2), above-mentioned classification is trained using the neural network in deep learning；
3), when real-time imaging or picture input, classification judgement is carried out by model, wherein the output probability of each classification
Between 0~1.0, a kind of classification for choosing maximum probability is that real-time imaging or picture are classified；
4) the default filter for, loading the classification automatically according to classification carries out filter processing to image or video frame；
5), image or video frame of the output after 4) step process.
2. a kind of method for judging automatically scenery type and carrying out image filters processing as described in claim 1, it is characterised in that:
Classification in the step 1) includes portrait, personage, natural land, indoor article, food, mechanical equipment, clothes vertical view
And/or it is other.
3. a kind of method for judging automatically scenery type and carrying out image filters processing as described in claim 1, it is characterised in that
The step 2) the following steps are included:
21), according to the classification in step 1), picture training sample is arranged respectively；
22) neural network, is constructed, the sample in step 21) is input to nerve net using the scheme of neural network classification
It is trained in network；
23), testing procedure 22) obtained model, until model has enough classification capacities.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810927477.2A CN109191371A (en) | 2018-08-15 | 2018-08-15 | A method of it judging automatically scenery type and carries out image filters processing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810927477.2A CN109191371A (en) | 2018-08-15 | 2018-08-15 | A method of it judging automatically scenery type and carries out image filters processing |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109191371A true CN109191371A (en) | 2019-01-11 |
Family
ID=64935864
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810927477.2A Withdrawn CN109191371A (en) | 2018-08-15 | 2018-08-15 | A method of it judging automatically scenery type and carries out image filters processing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109191371A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109903248A (en) * | 2019-02-20 | 2019-06-18 | 厦门美图之家科技有限公司 | A kind of method and image processing method generating automatic white balance model |
| CN109919867A (en) * | 2019-02-21 | 2019-06-21 | 成都品果科技有限公司 | A kind of filter method and device |
| CN110225221A (en) * | 2019-04-26 | 2019-09-10 | 广东虎彩影像有限公司 | A kind of automatic photo fix method and system |
| CN110225220A (en) * | 2019-04-26 | 2019-09-10 | 广东虎彩影像有限公司 | A kind of automatic photo fix system |
| CN110298405A (en) * | 2019-07-03 | 2019-10-01 | 北京字节跳动网络技术有限公司 | Classification recognition methods and device, storage medium and terminal |
| CN110599394A (en) * | 2019-09-12 | 2019-12-20 | 北京字节跳动网络技术有限公司 | Method and device for processing pictures in online presentation, storage medium and equipment |
| CN112118482A (en) * | 2020-09-17 | 2020-12-22 | 广州酷狗计算机科技有限公司 | Audio file playing method and device, terminal and storage medium |
-
2018
- 2018-08-15 CN CN201810927477.2A patent/CN109191371A/en not_active Withdrawn
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109903248A (en) * | 2019-02-20 | 2019-06-18 | 厦门美图之家科技有限公司 | A kind of method and image processing method generating automatic white balance model |
| CN109919867A (en) * | 2019-02-21 | 2019-06-21 | 成都品果科技有限公司 | A kind of filter method and device |
| CN110225221A (en) * | 2019-04-26 | 2019-09-10 | 广东虎彩影像有限公司 | A kind of automatic photo fix method and system |
| CN110225220A (en) * | 2019-04-26 | 2019-09-10 | 广东虎彩影像有限公司 | A kind of automatic photo fix system |
| CN110298405A (en) * | 2019-07-03 | 2019-10-01 | 北京字节跳动网络技术有限公司 | Classification recognition methods and device, storage medium and terminal |
| CN110599394A (en) * | 2019-09-12 | 2019-12-20 | 北京字节跳动网络技术有限公司 | Method and device for processing pictures in online presentation, storage medium and equipment |
| CN112118482A (en) * | 2020-09-17 | 2020-12-22 | 广州酷狗计算机科技有限公司 | Audio file playing method and device, terminal and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109191371A (en) | A method of it judging automatically scenery type and carries out image filters processing | |
| CN106161939B (en) | Photo shooting method and terminal | |
| CN105791705B (en) | Video anti-fluttering method, system and camera terminal suitable for mobile time-lapse photography | |
| CN103973977B (en) | Virtualization processing method, device and the electronic equipment of a kind of preview interface | |
| CN105243371B (en) | A kind of detection method, system and the camera terminal of face U.S. face degree | |
| CN109977739A (en) | Image processing method, device, storage medium and electronic equipment | |
| CN109191558A (en) | Image method for polishing and device | |
| CN108701359A (en) | Across the video frame tracking interest region with corresponding depth map | |
| CN110807759B (en) | Method and device for evaluating photo quality, electronic equipment and readable storage medium | |
| CN105631417A (en) | Video beautification system and method applied to Internet video live broadcast | |
| CN106683056A (en) | Airborne photoelectric infrared digital image processing method and apparatus thereof | |
| CN110324532A (en) | A kind of image weakening method, device, storage medium and electronic equipment | |
| CN107944420A (en) | The photo-irradiation treatment method and apparatus of facial image | |
| CN107172354A (en) | Method for processing video frequency, device, electronic equipment and storage medium | |
| CN110022430A (en) | Image weakening method, device, mobile terminal and computer readable storage medium | |
| CN109286758A (en) | A kind of generation method of high dynamic range images, mobile terminal and storage medium | |
| CN109063776A (en) | Image identifies network training method, device and image recognition methods and device again again | |
| CN110336945A (en) | A kind of intelligence assisted tomography patterning process and system | |
| CN111985281A (en) | Image generation model generation method and device and image generation method and device | |
| CN108184075A (en) | For generating the method and apparatus of image | |
| CN103888655B (en) | A kind of photographic method and electronic equipment | |
| CN108875751A (en) | Image processing method and device, the training method of neural network, storage medium | |
| CN112036209A (en) | Portrait photo processing method and terminal | |
| CN113781370A (en) | Image enhancement method and device and electronic equipment | |
| CN107133981B (en) | Image processing method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20190111 |
|
| WW01 | Invention patent application withdrawn after publication |