CN102646273A - Method for automatically generating and intelligently managing image collage based on importance detection - Google Patents
Method for automatically generating and intelligently managing image collage based on importance detection Download PDFInfo
- Publication number
- CN102646273A CN102646273A CN2011104461304A CN201110446130A CN102646273A CN 102646273 A CN102646273 A CN 102646273A CN 2011104461304 A CN2011104461304 A CN 2011104461304A CN 201110446130 A CN201110446130 A CN 201110446130A CN 102646273 A CN102646273 A CN 102646273A
- Authority
- CN
- China
- Prior art keywords
- image
- importance
- collage
- angle
- intelligent management
- 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
Abstract
The invention discloses a method for automatically generating and intelligently managing image collage based on importance detection. The method is characterized by comprising the following steps of: (1) performing image importance detection, namely extracting an important area of the image by employing the steps of color detection, depth of field analysis, composition analysis and the like; (2) performing intelligent arrangement, namely arranging the image on a canvas using an algorithm according to the obtained important image in the last step, wherein the important part is guaranteed to be not blocked in the arrangement process; and (3) diversifying the angle, namely supporting the angle diversity to have a positive effect to collage appearance through user-study. Therefore, an angle of each arranged image complying with a certain distribution is given. The generated collage has high information content, so that the integral information of a user is maximized; and meanwhile, the generated collage is basically filled with the canvas, so that the canvas has high utilization rate.
Description
Technical field
The present invention relates to automatic generation and the intelligent management of a kind of image collage that detects based on importance.
Background technology
Along with the development of digital camera, increasing digital photo album is widely used, such as the Picasa of Google and the Flickr of Yahoo; Because a large amount of images is arranged, just produced a problem thereupon, how to show effectively that these images have many products at present, as the mode of some simple image slideshows is provided.Yet these modes have not only been abandoned semantic information, and effect also is unsatisfactory.Based on above-mentioned, we have proposed automatic generation and the intelligent management of a kind of image collage that detects based on importance.
Summary of the invention
The technical matters that the present invention will solve is how to make the collage that generates comprise information as much as possible, though also the Global Information maximization that obtains of user and how to make that the utilization factor (image fills up) of painting canvas is high as far as possible.The present invention adopts following technical scheme for this reason, and this method may further comprise the steps:
(1) importance figure detects:
The importance zone of image is extracted in (1.1) to (1.3) set by step:
(11) color importance detects: according to the space distribution of color, and the relative importance between different colours, calculate the initial importance figure that reflects each pixel significance degree;
(1.2) depth of field analysis: input picture is carried out depth of field analysis,, then directly get into step (1.3) if image is long depth image; If image is shallow depth image, the focus of computed image according to the focal position, is revised the importance figure that a step obtains;
(1.3) composition analysis: image is carried out composition analysis,, further revise importance figure according to the architectural feature of image;
(2) intelligence is put: according to the importance figure that a last step obtains, use our algorithm on painting canvas, to put image, the process of putting guarantees that pith is not blocked:
(2.1) index selected digital image: for original image set I
m(1 ..., n), confirm a certain image through the selection of index, wherein n is the number of image;
(2.2) confirm the size of image on painting canvas: definition
is W wherein; H is the wide and high of painting canvas; W; H is the wide and high of image, and n is the number of image; Through the ratio of above-mentioned definition, the image that is placed on the painting canvas multiply by this ratio, confirm the size of image on painting canvas;
(2.3) put image: at first putting an image at the centre bit of painting canvas, is the center with this image then, to around diffusion put, whenever put an image, call and block algorithm and detect the importance zone and whether be blocked; If be blocked, adjust through the position and the size that change present image;
(3) angle diversity:
(3.1) angle is confirmed: the numbering of image (1 ..., n),, make the angle of image obey μ=± 10, δ=5 Gaussian distribution according to the parity of index as the index of unique definite image;
(3.2) angle rotation: the center with image serves as that fixed point is done rotation, the state set of definition image rotation 1,0, and 1} ,-1 presentation video is rotated counterclockwise, and the 0 expression anglec of rotation is that 0,1 presentation video turns clockwise;
(3.3) image angle self-optimizing: image changes the size of image slightly after step (3.2) rotates to an angle, and the blank around the image is covered;
(3.4) angle optimization between image: observe the position relation of any two images, if blank more between the lap of these two images and image rotated the angle that tries again of image wherein.
Automatic generation and the intelligent management of a kind of image collage that detects based on importance of the present invention, what the card of abandoning (collage) of generation comprised contains much information, and makes the Global Information maximization that the user obtains.The collage image that generates simultaneously fills up painting canvas basically, makes that the utilization factor of painting canvas is very high.
As further improvement of the present invention, the rectangle information representation is used in the importance of the image that step (1) is extracted zone.Described rectangle information comprises upper left apex coordinate, and the width of rectangle and height.Generally speaking, the image importance zone rectangle that we obtain can be less than normal, therefore, can suitably enlarge the information of this rectangular area, and exaggerated scale is 5% of image length and width.
Further improve again as of the present invention,, support the angle diversity the collage positive role that plays attractive in appearance through user-study.Therefore, each after we put is opened angle of obeying certain distributed of image.Owing to want image rotating, in the process of putting, we suitably enlarge the size in importance zone, after rotation, guarantee that still the actual importance zone is not blocked like this.For given angle, we let it obey the Gaussian distribution of expectation symmetry respectively according to the index parity, thereby have both guaranteed the angle diversity, and symmetrical beauty is arranged again.
Further improve again as of the present invention; The present invention also provides a kind of intelligent management pattern: through the collage of preceding step (3) generation; The satisfied place of people not to the utmost still, this is to need us through the mode of manual work to be made amendment in some goodish place.Specifically; Provide user interactions to revise the instrument of collage; Make the user can revise image easily, such as the information such as size, position, the anglec of rotation and tone of adjustment image, adjustment information highlights importance figure etc.; Make collage more meet user's demand, have more effect attractive in appearance.
As improvement further of the present invention, the present invention also provides a kind of intelligent browse mode: appreciate the whole structure of the collage of intelligence generation through series of operation is provided for the user.
Description of drawings
Fig. 1 is an overview flow chart of the present invention.
Fig. 2 is an image that detects the importance important area, and wherein the zone that comprises of rectangle message box is the importance zone.
Fig. 3 is the synoptic diagram that method provided by the present invention is made, and wherein (a) is the original image set, (b) is the collage after intelligence generation and the optimization.
Fig. 4 is the object lesson that method provided by the present invention is made, and wherein (a) is the original image set, (b) is the collage after intelligence generation and the optimization.
Fig. 5 is the location diagram of any two images.
Fig. 6 is the synoptic diagram to angle optimizing process between image.
Embodiment
At first define the abbreviation that to use in the next explanation: original image set I
i=1 ..., n}, importance atlas S
i=1 ..., n},
Fig. 1 is a basic flow sheet of the present invention, and invention is input with some row images, through a series of processing, obtains a width of cloth maximized the piecing together of certain appreciation effect that have that convey a message, and is elaborated in the face of each flow process of the present invention down:
1. importance figure detects
Here we import a series of original image set; Through steps such as color detection, depth of field analysis, composition analysis, extract the importance zone of image, represent with rectangle; For the back angle is rotated conveniently; Here enlarge the area in importance zone, and exaggerated scale is 5% of image length and width, thereby pith is not blocked after guaranteeing the image rotation.
Described color detection, depth of field analysis and composition analysis step are following:
(11) color importance detects: according to the space distribution of color, and the relative importance between different colours, calculate the initial importance figure that reflects each pixel sensitivity;
(1.2) depth of field analysis: input picture is carried out depth of field analysis,, then directly get into step (1.3) if image is long depth image; If image is shallow depth image, the focus of computed image according to the focal position, is revised the importance figure that a step obtains;
(1.3) composition analysis: image is carried out composition analysis,, further revise importance figure according to the architectural feature of image
2. intelligence is put
For original image set I
i(1 ..., n), index is unique confirms that through the selection of index, we can confirm a certain image.In order to make image that a suitable size arranged on painting canvas; We are definition
W wherein; H is the wide and high of painting canvas; W, h are the wide and high of image, and n is the number of image.Through the ratio of above-mentioned definition, the image that we will be placed on the painting canvas multiply by this ratio, just can confirm the size of image on painting canvas.
In the process of putting, at first put an image in the center (part at random) of painting canvas, be the center with this image then, to around diffusion put, whenever put an image, call and block algorithm and detect the importance area part and whether be blocked.If be blocked, information such as our position through changing present image, size are adjusted.The simple description of blocking algorithm is exactly whether the summit of judging present image importance zone drops on the zone of having put the image place, if then block, otherwise does not block.
3. angle diversity
(1) angle is confirmed: we the numbering of image (1 ..., n) as the index of unique definite image; According to the parity of index, make the angle of image obey μ=± 10, δ=5 Gaussian distribution; Make the angle of image can guarantee diversity like this, have to keep a kind of simple symmetrical beauty.
(2) angle rotation: the rotation of angle here is that the center with image is the rotation that fixed point is done, we define the image rotation state set 1,0,1} ,-1 presentation video is rotated counterclockwise, the 0 expression anglec of rotation is that 0,1 presentation video turns clockwise.Because when the anglec of rotation, image has been put on painting canvas and finished, the anglec of rotation will some blank can occur four angle parts after producing, and few as much as possible in order to guarantee these blank, we do the optimization of next 2 aspects.
(3) image angle self-optimizing:
Image is after above-mentioned rotating to an angle, and for a certain image itself, we will guarantee blank the least possible around the image, and we change the size of image slightly, make will around the blank covering.
(4) angle optimization between image:
With reference to Fig. 5, according to the state that the angle of front definition is rotated, the position relation that we observe any two images has following several kinds of situation:
For Fig. 5 (a) with the situation (b), blank and shield portions is in accepting scope; And for the situation among Fig. 5 (c); We will adjust by method shown in Figure 6: be about to image 2 along this picture centre certain angle that turns clockwise; Angle between image 1 and the image 2 is diminished, and process result makes lap and blank area between two images reduce like this.
4. intelligent management: for the collage of automatic generation; Always unavoidable some place is barely satisfactory; We provide some artificial modes to change the collage of front; The user can change the position, size, the anglec of rotation, figure layer, tone of image etc., and Fig. 3 is the design sketch of the final collage that generates.
5. intelligence is browsed: for the collage that the front generates, the mode that the user can like according to oneself is appreciated collage.
What should be understood that is: the foregoing description is just to explanation of the present invention, rather than limitation of the present invention, and any innovation and creation that do not exceed in the connotation scope of the present invention all fall within protection scope of the present invention.
Claims (7)
1. automatic generation and the intelligent management of an image collage who detects based on importance is characterized in that this method may further comprise the steps:
(1) importance detects:
The importance zone of image is extracted in (1.1) to (1.3) set by step:
(11) color importance detects: according to the space distribution of color, and the relative importance between different colours, calculate the initial importance figure that reflects each pixel sensitivity;
(1.2) depth of field analysis: input picture is carried out depth of field analysis,, then directly get into step (1.3) if image is long depth image; If image is shallow depth image, the focus of computed image according to the focal position, is revised the importance figure that a step obtains;
(1.3) composition analysis: image is carried out composition analysis,, further revise importance figure according to the architectural feature of image;
(2) intelligence is put: image
(2.1) index selected digital image: for original image set I
i(i=1 ..., n), confirm a certain image through the selection of index, wherein n is the number of image;
(2.2) confirm the size of image on painting canvas: definition
is W wherein; H is the wide and high of painting canvas; W; H is the wide and high of image, and n is the number of image; Through the ratio of above-mentioned definition, the image that is placed on the painting canvas multiply by this ratio, confirm the size of image on painting canvas;
(2.3) put image: at first putting an image at the centre bit of painting canvas, is the center with this image then, to around diffusion put, whenever put an image, call and block algorithm and detect the importance zone and whether be blocked; If be blocked, adjust through the position and the size that change present image;
(3) angle diversity: image
(3.1) angle is confirmed: the numbering of image (1 ..., n),, make the angle of image obey μ=± 10, δ=5 Gaussian distribution according to the parity of index as the index of unique definite image;
(3.2) angle rotation: the center with image serves as that fixed point is done rotation, the state set of definition image rotation 1,0, and 1} ,-1 presentation video is rotated counterclockwise, and the 0 expression anglec of rotation is that 0,1 presentation video turns clockwise;
(3.3) image angle self-optimizing: image changes the size of image slightly after step (3.2) rotates to an angle, and the blank around the image is covered;
(3.4) angle optimization between image: observe the position relation of any two images, if blank more between the lap of these two images and image rotated the angle that tries again of image wherein.
2. automatic generation and the intelligent management of a kind of image collage that detects based on significance map according to claim 1, the rectangle information representation is used in the importance zone of the image that it is characterized in that step (1) is extracted.
3. automatic generation and the intelligent management of a kind of image collage that detects based on significance map according to claim 2 is characterized in that described rectangle information comprises upper left apex coordinate, and the width of rectangle and height.
4. automatic generation and the intelligent management of a kind of image collage that detects based on significance map according to claim 2 is characterized in that described image importance zone rectangle is extended to 1-1.05 doubly.
5. automatic generation and the intelligent management of a kind of image collage that detects based on significance map according to claim 1; It is characterized in that the described algorithm that blocks of step (2.3) is meant: whether cycle criterion present image importance zone drops on the shared zone of other image; If; Then expression is blocked, otherwise, promptly do not block.
6. according to automatic generation and the intelligent management of the described a kind of image collage that detects based on significance map of right 1; It is characterized in that the card of abandoning for step (3) generation; The instrument that provides user interactions to revise collage,, through size, position, the anglec of rotation and the tone of manual type adjustment image; Adjustment figure layer information is further revised image.
7. according to automatic generation and the intelligent management of any one described a kind of image collage that detects based on significance map of right 1-6; It is characterized in that; The card of abandoning for step (3) generation; The mode that provides a kind of intelligence to browse, the mode that this intelligence is browsed comprises some simple user operation, the mode that the user selects oneself to like is browsed the card of abandoning.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2011104461304A CN102646273A (en) | 2012-03-29 | 2012-03-29 | Method for automatically generating and intelligently managing image collage based on importance detection |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2011104461304A CN102646273A (en) | 2012-03-29 | 2012-03-29 | Method for automatically generating and intelligently managing image collage based on importance detection |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN102646273A true CN102646273A (en) | 2012-08-22 |
Family
ID=46659085
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2011104461304A Pending CN102646273A (en) | 2012-03-29 | 2012-03-29 | Method for automatically generating and intelligently managing image collage based on importance detection |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102646273A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105376480A (en) * | 2014-08-19 | 2016-03-02 | 卡西欧计算机株式会社 | Image extracting apparatus and image extracting method |
| CN109829486A (en) * | 2019-01-11 | 2019-05-31 | 新华三技术有限公司 | Image processing method and device |
-
2012
- 2012-03-29 CN CN2011104461304A patent/CN102646273A/en active Pending
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105376480A (en) * | 2014-08-19 | 2016-03-02 | 卡西欧计算机株式会社 | Image extracting apparatus and image extracting method |
| CN105376480B (en) * | 2014-08-19 | 2018-08-03 | 卡西欧计算机株式会社 | Image acquiring apparatus, image extraction method and computer-readable storage medium |
| CN109829486A (en) * | 2019-01-11 | 2019-05-31 | 新华三技术有限公司 | Image processing method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US8849043B2 (en) | System for matching artistic attributes of secondary image and template to a primary image | |
| US8274523B2 (en) | Processing digital templates for image display | |
| US8212834B2 (en) | Artistic digital template for image display | |
| US8854395B2 (en) | Method for producing artistic image template designs | |
| US8890886B2 (en) | User interface with color themes based on input image data | |
| US8289340B2 (en) | Method of making an artistic digital template for image display | |
| US8849853B2 (en) | Method for matching artistic attributes of a template and secondary images to a primary image | |
| US20110029914A1 (en) | Apparatus for generating artistic image template designs | |
| CN101551904B (en) | Image synthesis method and apparatus based on mixed gradient field and mixed boundary condition | |
| US8345057B2 (en) | Context coordination for an artistic digital template for image display | |
| US20110029553A1 (en) | System for coordinating user images in an artistic design | |
| US20110029860A1 (en) | Artistic digital template for image display | |
| WO2017097143A1 (en) | Method and device for setting interface switch animation | |
| US20110029562A1 (en) | Coordinating user images in an artistic design | |
| US8332427B2 (en) | Method of generating artistic template designs | |
| CN103139478A (en) | Photographing method with real-time rendering of main image and secondary image | |
| CN109992266A (en) | A kind for the treatment of method and apparatus of interface element | |
| CN110706151A (en) | Video-oriented non-uniform style migration method | |
| CN107197171A (en) | A kind of digital photographing processing method for adding intelligence software light source | |
| CN102646273A (en) | Method for automatically generating and intelligently managing image collage based on importance detection | |
| CN106445284A (en) | Screen display method and device | |
| CN104978445B (en) | A kind of method and device of picture combination | |
| Myodo et al. | Automatic comic-like image layout system preserving image order and important regions | |
| CN109104627A (en) | Focus Background Generation Method, storage medium, equipment and the system of Android TV | |
| CN106446053A (en) | Method and device for displaying picture resources |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C12 | Rejection of a patent application after its publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20120822 |