CN111210434A - Image replacement method and system based on sky identification - Google Patents
Image replacement method and system based on sky identification Download PDFInfo
- Publication number
- CN111210434A CN111210434A CN201911319364.5A CN201911319364A CN111210434A CN 111210434 A CN111210434 A CN 111210434A CN 201911319364 A CN201911319364 A CN 201911319364A CN 111210434 A CN111210434 A CN 111210434A
- Authority
- CN
- China
- Prior art keywords
- sky
- information
- region
- area
- identification
- 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
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/20—Image enhancement or restoration by the use of local operators
- G06T5/30—Erosion or dilatation, e.g. thinning
-
- 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 by the use of more than one image, e.g. averaging, subtraction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20212—Image combination
- G06T2207/20221—Image fusion; Image merging
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30181—Earth observation
Abstract
The invention provides an image replacement method and system based on sky identification, comprising the following steps: sky identification step: acquiring sky region segmentation map information according to the original input image parameters; a sky material acquisition step: acquiring sky material template selection information according to system default parameters or user selection information; replacement fusion step: and replacing and fusing the sky area according to the sky material template selection information and the sky area segmentation map information to generate a final map after the sky is replaced, and acquiring sky replacement result information. The method uses a sky recognition algorithm to obtain a segmentation map of a sky area; the method can provide a large number of dazzling sky material templates for supply; according to the method and the device, the sky area can be replaced and fused according to the sky template map selected by the user, and the final map with the changed sky is generated.
Description
Technical Field
The invention relates to the field of image correction, in particular to an image replacement method and system based on sky identification.
Background
In daily photography, the sky is one of the most common backgrounds and appears in daily photos with high probability, but because the weather is uncontrollable, the beautiful sky is often an indispensable thing to be appeared in the photos.
Patent document CN106792147A discloses an image replacement method and device for realizing the effect of replacing an avatar in a video, so as to improve the experience of interest of a user. The method comprises the following steps: capturing a selection operation of a first face image in a current video interface, wherein the current video interface comprises at least one face image; acquiring a photo to be replaced, and identifying a second face image in the photo to be replaced; and replacing the first face image with the second face image. This patent is not well suited for use in the replacement of sky backgrounds.
Disclosure of Invention
In view of the defects in the prior art, the present invention aims to provide an image replacement method and system based on sky identification.
The invention provides an image replacement method based on sky identification, which comprises the following steps: sky identification step: acquiring sky region segmentation map information according to the original input image parameters; a sky material acquisition step: acquiring sky material template selection information according to system default parameters or user selection information; replacement fusion step: and replacing and fusing the sky area according to the sky material template selection information and the sky area segmentation map information to generate a final map after the sky is replaced, and acquiring sky replacement result information.
Preferably, the sky identifying step includes: a convolutional neural network model obtaining step: training a scene picture in a training set containing a sky area by using a convolutional neural network to obtain a convolutional neural network model for sky identification, and acquiring information of the convolutional neural network model for sky identification; a coarse-grained sky region acquisition step: and according to the model information of the sky recognition convolutional neural network, carrying out preliminary segmentation by using a trained model to obtain a coarse-grained sky region and obtain coarse-grained sky region information.
Preferably, the sky identifying step further comprises: and a step of obtaining the undetermined area: according to the coarse-grained sky region information, carrying out corrosion and expansion algorithms on the identified sky region to obtain a sky region, a determined non-sky region and a sky and non-sky boundary region, and acquiring information of a region to be determined; the information of the undetermined area is matched with the information of the sky and the non-sky boundary area.
Preferably, the sky identifying step further comprises: dividing the region to be determined: extracting a gradient information graph from an original image by using a sobel operator according to information of an undetermined area, setting a threshold value, scanning a boundary area from top to bottom according to columns to obtain a boundary, and preliminarily dividing the undetermined area into a sky part and a non-sky part; obtaining an optimal segmentation line: an energy function is defined for calculating the rationality of the partitioning. The above segmentation can be regarded as a two-classification problem, and in order to make the segmentation more reasonable, it is necessary to make the differences within the classes as small as possible, i.e. the covariance as small as possible. We therefore define the energy function principle as follows: calculating the minimum value of covariance of sky and ground areas to maximize an energy function, and taking a boundary line with the maximum energy function as an optimal dividing line; judging the Mahalanobis distance: and scanning each column of pixels in the undetermined area, and judging whether the pixel belongs to a sky pixel or not through the Mahalanobis distance between each column of pixels and the determined sky area in the previous algorithm, so that the effect of the sky boundary line can be further improved.
Preferably, the sky material acquiring step: visual effect display step: and acquiring visual effect display information according to default parameters of the system or user selection information, so that the user can visually see the effect conveniently.
The invention provides an image replacement system based on sky identification, which comprises: the sky identification module: acquiring sky region segmentation map information according to the original input image parameters; sky material acquisition module: acquiring sky material template selection information according to system default parameters or user selection information; replacing the fusion module: and replacing and fusing the sky area according to the sky material template selection information and the sky area segmentation map information to generate a final map after the sky is replaced, and acquiring sky replacement result information.
Preferably, the sky identification module includes: a convolutional neural network model acquisition module: training a scene picture in a training set containing a sky area by using a convolutional neural network to obtain a convolutional neural network model for sky identification, and acquiring information of the convolutional neural network model for sky identification; coarse-grained sky region acquisition module: and according to the model information of the sky recognition convolutional neural network, carrying out preliminary segmentation by using a trained model to obtain a coarse-grained sky region and obtain coarse-grained sky region information.
Preferably, the sky identification module further comprises: undetermined area acquisition module: according to the coarse-grained sky region information, carrying out corrosion and expansion algorithms on the identified sky region to obtain a sky region, a determined non-sky region and a sky and non-sky boundary region, and acquiring information of a region to be determined; the information of the undetermined area is matched with the information of the sky and the non-sky boundary area.
Preferably, the sky identification module further comprises: the pending area division module: extracting a gradient information graph from an original image by using a sobel operator according to information of an undetermined area, setting a threshold value, scanning a boundary area from top to bottom according to columns to obtain a boundary, and preliminarily dividing the undetermined area into a sky part and a non-sky part; an optimal segmentation line acquisition module: an energy function is defined for calculating the rationality of the partitioning. The above segmentation can be regarded as a two-classification problem, and in order to make the segmentation more reasonable, it is necessary to make the differences within the classes as small as possible, i.e. the covariance as small as possible. We therefore define the energy function principle as follows: calculating the minimum value of covariance of sky and ground areas to maximize an energy function, and taking a boundary line with the maximum energy function as an optimal dividing line; the Mahalanobis distance judgment module: and scanning each column of pixels in the undetermined area, and judging whether the pixel belongs to a sky pixel or not through the Mahalanobis distance between each column of pixels and the determined sky area in the previous algorithm, so that the effect of the sky boundary line can be further improved.
Preferably, the sky material acquiring module: the visual effect display module: and acquiring visual effect display information according to default parameters of the system or user selection information, so that the user can visually see the effect conveniently.
Compared with the prior art, the invention has the following beneficial effects:
1. the method uses a sky recognition algorithm to obtain a segmentation map of a sky area;
2. the method can provide a large number of dazzling sky material templates for supply;
3. according to the method and the device, the sky area can be replaced and fused according to the sky template map selected by the user, and the final map with the changed sky is generated.
Drawings
Other features, objects and advantages of the invention will become more apparent upon reading of the detailed description of non-limiting embodiments with reference to the following drawings:
FIG. 1 is a schematic flow chart of the present invention.
FIG. 2 is a system framework diagram of the present invention.
Detailed Description
The present invention will be described in detail with reference to specific examples. The following examples will assist those skilled in the art in further understanding the invention, but are not intended to limit the invention in any way. It should be noted that it would be obvious to those skilled in the art that various changes and modifications can be made without departing from the spirit of the invention. All falling within the scope of the present invention.
As shown in fig. 1 and 2, an image replacement method based on sky recognition according to the present invention includes: sky identification step: acquiring sky region segmentation map information according to the original input image parameters; a sky material acquisition step: acquiring sky material template selection information according to system default parameters or user selection information; replacement fusion step: and replacing and fusing the sky area according to the sky material template selection information and the sky area segmentation map information to generate a final map after the sky is replaced, and acquiring sky replacement result information.
Preferably, the sky identifying step includes: a convolutional neural network model obtaining step: training a scene picture in a training set containing a sky area by using a convolutional neural network to obtain a convolutional neural network model for sky identification, and acquiring information of the convolutional neural network model for sky identification; a coarse-grained sky region acquisition step: and according to the model information of the sky recognition convolutional neural network, carrying out preliminary segmentation by using a trained model to obtain a coarse-grained sky region and obtain coarse-grained sky region information.
Preferably, the sky identifying step further comprises: and a step of obtaining the undetermined area: according to the coarse-grained sky region information, carrying out corrosion and expansion algorithms on the identified sky region to obtain a sky region, a determined non-sky region and a sky and non-sky boundary region, and acquiring information of a region to be determined; the information of the undetermined area is matched with the information of the sky and the non-sky boundary area.
Preferably, the sky identifying step further comprises: dividing the region to be determined: extracting a gradient information graph from an original image by using a sobel operator according to information of an undetermined area, setting a threshold value, scanning a boundary area from top to bottom according to columns to obtain a boundary, and preliminarily dividing the undetermined area into a sky part and a non-sky part; obtaining an optimal segmentation line: an energy function is defined for calculating the rationality of the partitioning. The above segmentation can be regarded as a two-classification problem, and in order to make the segmentation more reasonable, it is necessary to make the differences within the classes as small as possible, i.e. the covariance as small as possible. We therefore define the energy function principle as follows: calculating the minimum value of covariance of sky and ground areas to maximize an energy function, and taking a boundary line with the maximum energy function as an optimal dividing line; judging the Mahalanobis distance: and scanning each column of pixels in the undetermined area, and judging whether the pixel belongs to a sky pixel or not through the Mahalanobis distance between each column of pixels and the determined sky area in the previous algorithm, so that the effect of the sky boundary line can be further improved.
Preferably, the sky material acquiring step: visual effect display step: and acquiring visual effect display information according to default parameters of the system or user selection information, so that the user can visually see the effect conveniently.
Specifically, in one embodiment, an image replacement method based on sky identification includes:
sky identification algorithm step:
1. and training the scene pictures in the training set containing the sky area by using a convolutional neural network to obtain a convolutional neural network model for sky identification.
2. And carrying out preliminary segmentation on the picture to be detected by using a trained model to obtain a coarse-grained sky region.
3. Carrying out erosion and expansion algorithm on the identified sky area to obtain three areas: the area after corrosion is regarded as a determined sky area, the content outside the area after expansion is regarded as a determined non-sky area, and the area between the area and the area is regarded as a boundary area of sky and non-sky, namely the area to be determined.
4. And further refining the region to be determined. We use the gradient information in the boundary region to precisely demarcate the boundary between sky and non-sky. The method specifically comprises the following steps:
a gradient information graph is extracted from an original image by using a sobel operator, a threshold value is set, a boundary area is scanned from top to bottom according to columns to obtain a boundary, and a region to be determined is preliminarily divided into a sky part and a non-sky part.
An energy function is defined for calculating the rationality of the partitioning. The above segmentation can be regarded as a two-classification problem, and in order to make the segmentation more reasonable, it is necessary to make the differences within the classes as small as possible, i.e. the covariance as small as possible. We therefore define the energy function principle as follows: i.e. the minimum of the covariance of the sky and ground areas is calculated so that the energy function is maximized. And taking the boundary line with the maximum energy function as the optimal dividing line.
According to the results of practical tests, the effect is improved by the following modes: and scanning each column of pixels in the undetermined area, and judging whether the pixel belongs to a sky pixel or not through the Mahalanobis distance between each column of pixels and the determined sky area in the previous algorithm, so that the effect of the sky boundary line can be further improved.
The user can select mass sky dazzling materials and can directly apply preview effect.
The sky region replacement and fusion step comprises the following steps:
1. the user-selected material is applied to the previously detected sky region.
2. And fusing the replaced area and the original image by adopting a common fusion algorithm (weighted fusion, Poisson fusion and the like), so that the whole image looks more natural.
One skilled in the art may understand the sky recognition-based image replacement method provided by the present invention as an embodiment of the sky recognition-based image replacement system provided by the present invention. That is, the sky identification-based image replacement system may be implemented by executing a flow of steps of the sky identification-based image replacement method.
The invention provides an image replacement system based on sky identification, which comprises: the sky identification module: acquiring sky region segmentation map information according to the original input image parameters; sky material acquisition module: acquiring sky material template selection information according to system default parameters or user selection information; replacing the fusion module: and replacing and fusing the sky area according to the sky material template selection information and the sky area segmentation map information to generate a final map after the sky is replaced, and acquiring sky replacement result information.
Preferably, the sky identification module includes: a convolutional neural network model acquisition module: training a scene picture in a training set containing a sky area by using a convolutional neural network to obtain a convolutional neural network model for sky identification, and acquiring information of the convolutional neural network model for sky identification; coarse-grained sky region acquisition module: and according to the model information of the sky recognition convolutional neural network, carrying out preliminary segmentation by using a trained model to obtain a coarse-grained sky region and obtain coarse-grained sky region information.
Preferably, the sky identification module further comprises: undetermined area acquisition module: according to the coarse-grained sky region information, carrying out corrosion and expansion algorithms on the identified sky region to obtain a sky region, a determined non-sky region and a sky and non-sky boundary region, and acquiring information of a region to be determined; the information of the undetermined area is matched with the information of the sky and the non-sky boundary area.
Preferably, the sky identification module further comprises: the pending area division module: extracting a gradient information graph from an original image by using a sobel operator according to information of an undetermined area, setting a threshold value, scanning a boundary area from top to bottom according to columns to obtain a boundary, and preliminarily dividing the undetermined area into a sky part and a non-sky part; an optimal segmentation line acquisition module: an energy function is defined for calculating the rationality of the partitioning. The above segmentation can be regarded as a two-classification problem, and in order to make the segmentation more reasonable, it is necessary to make the differences within the classes as small as possible, i.e. the covariance as small as possible. We therefore define the energy function principle as follows: calculating the minimum value of covariance of sky and ground areas to maximize an energy function, and taking a boundary line with the maximum energy function as an optimal dividing line; the Mahalanobis distance judgment module: and scanning each column of pixels in the undetermined area, and judging whether the pixel belongs to a sky pixel or not through the Mahalanobis distance between each column of pixels and the determined sky area in the previous algorithm, so that the effect of the sky boundary line can be further improved.
Preferably, the sky material acquiring module: the visual effect display module: and acquiring visual effect display information according to default parameters of the system or user selection information, so that the user can visually see the effect conveniently.
The method uses a sky recognition algorithm to obtain a segmentation map of a sky area; the method can provide a large number of dazzling sky material templates for supply; according to the method and the device, the sky area can be replaced and fused according to the sky template map selected by the user, and the final map with the changed sky is generated.
Those skilled in the art will appreciate that, in addition to implementing the system and its various devices, units provided by the present invention as pure computer readable program code, the system and its various devices, units provided by the present invention can be fully enabled to implement the same functions by logically programming the method steps in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers, embedded microcontrollers and the like. Therefore, the system and various devices, units and units thereof provided by the invention can be regarded as a hardware component, and the devices, units and units included in the system for realizing various functions can also be regarded as structures in the hardware component; means, elements, units for performing various functions may also be regarded as structures within both software and hardware components for performing the method.
The foregoing description of specific embodiments of the present invention has been presented. It is to be understood that the present invention is not limited to the specific embodiments described above, and that various changes or modifications may be made by one skilled in the art within the scope of the appended claims without departing from the spirit of the invention. The embodiments and features of the embodiments of the present application may be combined with each other arbitrarily without conflict.
Claims (10)
1. An image replacement method based on sky recognition, comprising:
sky identification step: acquiring sky region segmentation map information according to the original input image parameters;
a sky material acquisition step: acquiring sky material template selection information according to system default parameters or user selection information;
replacement fusion step: and replacing and fusing the sky area according to the sky material template selection information and the sky area segmentation map information to generate a final map after the sky is replaced, and acquiring sky replacement result information.
2. The sky identification based image replacement method of claim 1, wherein said sky identification step comprises:
a convolutional neural network model obtaining step: training a scene picture in a training set containing a sky area by using a convolutional neural network to obtain a convolutional neural network model for sky identification, and acquiring information of the convolutional neural network model for sky identification;
a coarse-grained sky region acquisition step: and according to the model information of the sky recognition convolutional neural network, carrying out preliminary segmentation by using a trained model to obtain a coarse-grained sky region and obtain coarse-grained sky region information.
3. The sky identification based image replacement method of claim 2, wherein said sky identification step further comprises:
and a step of obtaining the undetermined area: according to the coarse-grained sky region information, carrying out corrosion and expansion algorithms on the identified sky region to obtain a sky region, a determined non-sky region and a sky and non-sky boundary region, and acquiring information of a region to be determined;
the information of the undetermined area is matched with the information of the sky and the non-sky boundary area.
4. The sky identification based image replacement method of claim 3, wherein said sky identification step further comprises:
dividing the region to be determined: extracting a gradient information graph from an original image by using a sobel operator according to information of an undetermined area, setting a threshold value, scanning a boundary area from top to bottom according to columns to obtain a boundary, and preliminarily dividing the undetermined area into a sky part and a non-sky part;
obtaining an optimal segmentation line: defining an energy function, calculating the minimum value of covariance of sky and ground areas to make the energy function maximum, and taking the boundary line with the maximum energy function as an optimal dividing line;
judging the Mahalanobis distance: and scanning each column of pixels in the region to be determined, and judging whether the pixel belongs to the sky pixel or not according to the Mahalanobis distance between each column of pixels and the determined sky region in the previous algorithm.
5. The sky recognition-based image replacement method of claim 1, wherein the sky material obtaining step:
visual effect display step: and acquiring visual effect display information according to default parameters of the system or information selected by a user.
6. An image replacement system based on sky identification, comprising:
the sky identification module: acquiring sky region segmentation map information according to the original input image parameters;
sky material acquisition module: acquiring sky material template selection information according to system default parameters or user selection information;
replacing the fusion module: and replacing and fusing the sky area according to the sky material template selection information and the sky area segmentation map information to generate a final map after the sky is replaced, and acquiring sky replacement result information.
7. The sky identification based image replacement system of claim 6, wherein the sky identification module comprises:
a convolutional neural network model acquisition module: training a scene picture in a training set containing a sky area by using a convolutional neural network to obtain a convolutional neural network model for sky identification, and acquiring information of the convolutional neural network model for sky identification;
coarse-grained sky region acquisition module: and according to the model information of the sky recognition convolutional neural network, carrying out preliminary segmentation by using a trained model to obtain a coarse-grained sky region and obtain coarse-grained sky region information.
8. The sky identification based image replacement system of claim 7, wherein the sky identification module further comprises:
undetermined area acquisition module: according to the coarse-grained sky region information, carrying out corrosion and expansion algorithms on the identified sky region to obtain a sky region, a determined non-sky region and a sky and non-sky boundary region, and acquiring information of a region to be determined;
the information of the undetermined area is matched with the information of the sky and the non-sky boundary area.
9. The sky identification based image replacement system of claim 8, wherein the sky identification module further comprises:
the pending area division module: extracting a gradient information graph from an original image by using a sobel operator according to information of an undetermined area, setting a threshold value, scanning a boundary area from top to bottom according to columns to obtain a boundary, and preliminarily dividing the undetermined area into a sky part and a non-sky part;
an optimal segmentation line acquisition module: defining an energy function, calculating the minimum value of covariance of sky and ground areas to make the energy function maximum, and taking the boundary line with the maximum energy function as an optimal dividing line;
the Mahalanobis distance judgment module: and scanning each column of pixels in the region to be determined, and judging whether the pixel belongs to the sky pixel or not according to the Mahalanobis distance between each column of pixels and the determined sky region in the previous algorithm.
10. The sky identification based image replacement system of claim 6, wherein the sky material acquisition module:
the visual effect display module: and acquiring visual effect display information according to default parameters of the system or information selected by a user.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911319364.5A CN111210434A (en) | 2019-12-19 | 2019-12-19 | Image replacement method and system based on sky identification |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911319364.5A CN111210434A (en) | 2019-12-19 | 2019-12-19 | Image replacement method and system based on sky identification |
Publications (1)
Publication Number | Publication Date |
---|---|
CN111210434A true CN111210434A (en) | 2020-05-29 |
Family
ID=70786247
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201911319364.5A Pending CN111210434A (en) | 2019-12-19 | 2019-12-19 | Image replacement method and system based on sky identification |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111210434A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112598679A (en) * | 2020-12-16 | 2021-04-02 | 中山大学 | Sky region segmentation method and device |
CN112613372A (en) * | 2020-12-16 | 2021-04-06 | 中山大学 | Outdoor environment visual inertia SLAM method and device |
WO2022194079A1 (en) * | 2021-03-19 | 2022-09-22 | 影石创新科技股份有限公司 | Sky region segmentation method and apparatus, computer device, and storage medium |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105631880A (en) * | 2015-12-31 | 2016-06-01 | 百度在线网络技术(北京)有限公司 | Lane line segmentation method and apparatus |
CN105869174A (en) * | 2016-04-20 | 2016-08-17 | 中国航空工业集团公司洛阳电光设备研究所 | Sky scene image segmentation method |
CN107025457A (en) * | 2017-03-29 | 2017-08-08 | 腾讯科技(深圳)有限公司 | A kind of image processing method and device |
WO2018052587A1 (en) * | 2016-09-14 | 2018-03-22 | Konica Minolta Laboratory U.S.A., Inc. | Method and system for cell image segmentation using multi-stage convolutional neural networks |
CN108961303A (en) * | 2018-07-23 | 2018-12-07 | 北京旷视科技有限公司 | A kind of image processing method, device, electronic equipment and computer-readable medium |
US20190122041A1 (en) * | 2017-06-06 | 2019-04-25 | Midea Group Co., Ltd. | Coarse-to-fine hand detection method using deep neural network |
CN110136143A (en) * | 2019-05-16 | 2019-08-16 | 河南大学 | Geneva based on ADMM algorithm multiresolution remote sensing image segmentation method off field |
CN110400323A (en) * | 2019-07-30 | 2019-11-01 | 上海艾麒信息科技有限公司 | It is a kind of to scratch drawing system, method and device automatically |
-
2019
- 2019-12-19 CN CN201911319364.5A patent/CN111210434A/en active Pending
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105631880A (en) * | 2015-12-31 | 2016-06-01 | 百度在线网络技术(北京)有限公司 | Lane line segmentation method and apparatus |
CN105869174A (en) * | 2016-04-20 | 2016-08-17 | 中国航空工业集团公司洛阳电光设备研究所 | Sky scene image segmentation method |
WO2018052587A1 (en) * | 2016-09-14 | 2018-03-22 | Konica Minolta Laboratory U.S.A., Inc. | Method and system for cell image segmentation using multi-stage convolutional neural networks |
CN107025457A (en) * | 2017-03-29 | 2017-08-08 | 腾讯科技(深圳)有限公司 | A kind of image processing method and device |
US20190122041A1 (en) * | 2017-06-06 | 2019-04-25 | Midea Group Co., Ltd. | Coarse-to-fine hand detection method using deep neural network |
CN108961303A (en) * | 2018-07-23 | 2018-12-07 | 北京旷视科技有限公司 | A kind of image processing method, device, electronic equipment and computer-readable medium |
CN110136143A (en) * | 2019-05-16 | 2019-08-16 | 河南大学 | Geneva based on ADMM algorithm multiresolution remote sensing image segmentation method off field |
CN110400323A (en) * | 2019-07-30 | 2019-11-01 | 上海艾麒信息科技有限公司 | It is a kind of to scratch drawing system, method and device automatically |
Non-Patent Citations (1)
Title |
---|
WEI WANG ET AL.: "PolSAR image segmentation based on hierarchical region merging and segment refinement with WMRF model", 《2017 IEEE INTERNATIONAL GEOSCIENCE AND REMOTE SENSING SYMPOSIUM》 * |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112598679A (en) * | 2020-12-16 | 2021-04-02 | 中山大学 | Sky region segmentation method and device |
CN112613372A (en) * | 2020-12-16 | 2021-04-06 | 中山大学 | Outdoor environment visual inertia SLAM method and device |
CN112598679B (en) * | 2020-12-16 | 2024-03-29 | 中山大学 | Sky area segmentation method and device |
CN112613372B (en) * | 2020-12-16 | 2024-04-02 | 中山大学 | Outdoor environment visual inertia SLAM method and device |
WO2022194079A1 (en) * | 2021-03-19 | 2022-09-22 | 影石创新科技股份有限公司 | Sky region segmentation method and apparatus, computer device, and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US11551338B2 (en) | Intelligent mixing and replacing of persons in group portraits | |
CN111210434A (en) | Image replacement method and system based on sky identification | |
CN103077388A (en) | Rapid text scanning method oriented to portable computing equipment | |
CN103098078A (en) | Smile detection systems and methods | |
CN106202086B (en) | Picture processing and obtaining method, device and system | |
CN105825478A (en) | Structure analysis method for recovering missing structures in an image after object removal | |
CN106203454A (en) | The method and device that certificate format is analyzed | |
CN105516590B (en) | A kind of image processing method and device | |
JP2009268085A (en) | Image trimming device and program | |
CN111145135A (en) | Image descrambling processing method, device, equipment and storage medium | |
CN115035580A (en) | Figure digital twinning construction method and system | |
CN115967823A (en) | Video cover generation method and device, electronic equipment and readable medium | |
CN104036272A (en) | Text detection method and electronic device | |
CN113225451A (en) | Image processing method and device and electronic equipment | |
CN112307923A (en) | Partitioned expression migration method and system | |
CN104504712A (en) | Picture processing method and device | |
CN112822393B (en) | Image processing method and device and electronic equipment | |
CN103530879B (en) | Pedestrian's color extraction method under special scenes | |
JP2007156741A (en) | Character extraction method, character extraction device, and program | |
CN113191942A (en) | Method for generating image, method for training human detection model, program, and device | |
JP2009277001A (en) | Image processor | |
CN113012085A (en) | Image processing method and device | |
JP2013122723A (en) | Detection device and detection method | |
CN113766147A (en) | Method for embedding image in video, and method and device for acquiring plane prediction model | |
CN113255421A (en) | Image detection method, system, device and medium |
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 |
Application publication date: 20200529 |
|
RJ01 | Rejection of invention patent application after publication |