KR101864454B1 - Apparatus and method for composing images in an image processing device - Google Patents

Abstract

The present invention relates to an apparatus and method for composing an image in an image processing apparatus, in which it is confirmed whether there is at least one object included in the images in an overlapping area between a plurality of images, If present, generates a Gaussian curve for the object, determines an alpha mask for the overlap region, and synthesizes the images based on the alpha mask.

Description

[0001] APPARATUS AND METHOD FOR COMPOSING IMAGES IN AN IMAGE PROCESSING DEVICE [0002]

The present invention relates to an image processing apparatus, and more particularly, to an apparatus and method for synthesizing an image in an image processing apparatus.

The plurality of images may be combined together to form a composite image. Some objects may be in an image that can be distinguished by foreground or background. These plurality of images may overlap each other to obtain a satisfactory composite image. This composite image may be synthesized using at least one of the available synthesis techniques such as alpha blending.

The alpha blending technique is a technique for combining source images by generating an image synthesized using the weights of corresponding pixels in a plurality of source images having foreground image opaque graphic effects. The alpha / transparency value of the alpha mask has a range of '0' to '1', '0' means a transparent pixel, and '1' means opaque pixel. The alpha value aids in the synthesis of foreground and background pixels, and the distribution for the foreground and background pixels is determined by the alpha value.

However, the alpha blending technique does not take advantage of the correlation between the content information of the synthesized images, which can reduce the visual appeal of the composite image by compositing the objects with the background / other images. The combination of these objects with the background / other images may make the objects partially or completely blurred, so that the content needs to be adjusted to make it easier to read. In the alpha blending technique, image frequencies are affected by interference effects. And the alpha blending technique induces visual ambiguity by inducing fading in the composite image. Sometimes it can be difficult to identify which layer belongs to which layer in an alpha blended image.

Also, in alpha blending techniques, alpha mask generation results in visible seam and ghosting effects. Visible shims may appear when the overlap window is too small or the two images to be composited are very distinct, whereas the ghosting effect may appear when the overlap window is too large.

Currently, blending techniques rely on creating optimal window sizes to avoid shim and ghosting effects. And blending techniques do not address the ghosting problem when the overlap window size can be fixed by an external module and the overlap window size can not be changed during composition.

The present invention proposes an apparatus and method for efficiently superimposing images without losing objects of interest when compositing images.

The present invention proposes an apparatus and method for efficiently superimposing fixed size images of overlapping areas.

According to an aspect of the present invention, there is provided a method of compositing an image in an image processing apparatus, comprising: checking whether at least one object included in the images exists in an overlapping region between a plurality of images And generating a Gauss curve for the object if the object exists, determining an alpha mask for the overlap region, and compositing the images based on the alpha mask.

According to an aspect of the present invention, there is provided an apparatus for compositing an image in an image processing apparatus, the apparatus comprising: a display unit; And if the object exists, generating an Gaussian curve for the object, determining an alpha mask for the overlap region, and synthesizing the images based on the alpha mask, .

The present invention has the effect of efficiently superimposing images without losing objects of interest when compositing images.

The present invention has the effect of efficiently superimposing fixed size images of overlapping areas.

1 is a block diagram of an image processing apparatus according to an embodiment of the present invention;
Figure 2 illustrates different images in accordance with embodiments of the present invention;
Figure 3 shows a composite image in accordance with embodiments of the present invention,
4 is a flowchart illustrating an image combining process performed by an image processing apparatus according to an embodiment of the present invention;
5 is a detailed view of the location of an object in an overlapping area according to embodiments of the present invention,
Figure 6 is a graph showing a Gaussian curve comprising an object in accordance with embodiments of the present invention;
Figure 7 is a graph showing the slope of a simple linear ramp according to embodiments of the present invention,
Figure 8 illustrates a plurality of objects in an overlap region according to embodiments of the present invention;
Figure 9 illustrates a composite guide profile constructed from individual curves of objects in accordance with embodiments of the present invention;
10 is a diagram showing an image synthesized by alpha mask generation according to embodiments of the present invention.

Embodiments, various features, and advantages of the present invention will be described in detail with reference to the accompanying drawings and the non-limiting embodiments described in the following description. The descriptions of well-known components and processing techniques are omitted when it is determined that the gist of the embodiments of the present invention may be blurred. It will be apparent to those skilled in the art that the examples used in the present invention are provided for a better understanding of the present invention and that the present invention can be practiced without these specific details. Accordingly, the above examples should not be construed as limiting the scope of the invention.

The present invention provides an apparatus and method for guiding an alpha mask generation based profile. 1 to 10 show preferred embodiments of the present invention.

Throughout the specification, words such as visual content and images are used interchangeably. The image may be a digital image that can represent raster images / alpha map images. And the images may be captured by a device using analog media such as a digital camera, or similar devices and / or photographs, photographic film, or prints by an image scanner or similar device.

Throughout the specification, words such as alpha masks, alpha channels, and alpha matrices are used analogously and interchangeably to refer to a matrix or array that constitutes alpha values for a superimposed area.

1 is a block diagram of an image processing apparatus according to an embodiment of the present invention.

1, an image processing apparatus includes an image synthesizing apparatus 102, an input apparatus 101, and an output apparatus 107. [ Here, the image synthesizing apparatus 102 may include a digital image decoder 104, a memory 105, an image processing unit 106, and an alpha mask generating unit 103. And memory 105 may store images or other visual content. And the image decoder 104 may decode the received images. The image decoder 104 may then decode the pixel values of the images. And the alpha mask generation unit 103 can generate alpha mask array data that can be programmed and used by the image processing unit 106. [ And the image processing unit 106 can analyze and process the received images. The image processing unit 106 may synthesize images with each other. And the synthesized image can be displayed on the output device 107. [ And the synthesized image may be output by an output device such as a screen / monitor, personal computer, PDA, mobile device, digital picture frame, and printer.

And the received image may be received from a digital camera (not shown) that allows users to capture images and store them in memory / database 105 in digital format. Here, the digital camera may be a stand-alone device, or it may be part of another device such as a mobile device, a portable personal terminal. And the images to be composited are transferred to the image synthesizing apparatus 102 through an input device 101 such as a mobile device, a camera, a database, a memory 105, a PDA, a scanner, compact discs (CDs) Can be provided.

And the image compositing device 102 may interact with at least the remote system on a network such as the Internet. Here, the network may be a wired or wireless communication network. And the remote system may be configured to provide image services and goods on the network. And the remote system may include at least one image rendering facility and a data center. The image compositing device 102 may then receive images from the data center of the remote system on the network. Here, the images can be delivered over the Internet via e-mail from the remote system to the image composing device 102 for compositing. The image synthesizer 102 may then process the images and synthesize them. Here, the synthesized images can be delivered to the remote system and displayed on the remote system.

FIG. 2 is a view showing different images according to embodiments of the present invention, and FIG. 3 is a view showing a composite image according to embodiments of the present invention.

Referring to Figures 2 and 3, Image 1 201 and Image 2 202 represent two images that need to be combined. The image synthesis shows a case where images are combined with each other in the overlapping area 301 to obtain a single image by synthesizing two images that are different from each other and / or different from each other. Here, the image 2 202 includes the object 203.

Images 1 (201) and 202 (2) in FIG. 2 may be combined with each other on the overlapping area 301 to obtain a composite image. If image 1 201 and image 2 202 are overlapped, image 1 201 may be the background image of the synthesized image and image 2 202 may be the foreground image of the synthesized image have. At this time, the object 203 on the image 2 202 may be located in the overlap area (area 1) 301. And the area where the image 1 (201) overlaps with the image 2 (202) may be the area 2 (302). At this time, the object 203 may be partially present in the overlapping area 301. For example, the object 203 may be a face, and a portion of the face may be in the overlap region 301. [ The width 'w _offset ' and the height 'h _offset ' of the overlap area 301 are determined. A portion of the object 203 located in the overlap region 301 is detected and an alpha blending scheme is performed on a portion of the object 203 to synthesize the object 203 without fading on the image 1 201 .

FIG. 4 is a flowchart illustrating an image combining process performed by an image processing apparatus according to an embodiment of the present invention. FIG.

Referring to FIG. 4, in step 401, the image synthesizing apparatus 102 receives images to be synthesized from the input apparatus 101, and then proceeds to step 402. In step 402, the image synthesizing apparatus 102 analyzes the received images according to conditions specified by the user or predetermined conditions, and then proceeds to step 403. At this time, the external module can provide information related to the overlap area, such as the height and width of the overlap area, to the image synthesizer 102 through the input device 101. In step 403, the image synthesizing apparatus 102 analyzes the received images to identify at least one object in the image to be synthesized, and then proceeds to step 404. In step 404, the image synthesizing apparatus 102 confirms whether or not the identified object is detected in the overlap area 301. [ If it is determined that the identified object is detected in the overlap area 301, the image synthesizing apparatus 102 proceeds to step 405. [ In step 405, the image synthesizer 102 generates a Gaussian curve for the identified object, and then proceeds to step 406. In step 406, In step 406, the image synthesizing apparatus 102 generates an alpha mask for the object 203 existing in the overlap area 301 using the Gaussian profile, and then proceeds to step 407. In step 407, the image synthesizer 102 synthesizes the images along a Gaussian curve using a well-known synthesis method such as an alpha blending method.

As a result of checking, if the identified object is not detected in the overlapping area 301, the image synthesizing apparatus 102 proceeds to step 406. [ In step 406, the image synthesizer 102 generates an alpha mask for the object 203, and then proceeds to step 407. In step 407, the image synthesizing apparatus 102 synthesizes the images using a known blending method.

Steps 401 to 407 may be performed in an existing order according to various embodiments, or a sequence or specific steps different from the existing order may be performed at the same time. In some embodiments, some of the steps shown in FIG. 4 may be omitted.

5 is a detailed view showing the position of an object in the overlapping area according to the embodiments of the present invention.

Referring to FIG. 5, a portion of the object 203 may be in the overlap region 501. Here, the width of the overlapping area 501 and the width of the object 203 are determined as 'w _offset ' and 'w _object ', respectively. The width from the beginning of the overlap area 301 to the object 203 is determined as 'x _offset '. The height of the overlapping area 301 and the height of the object 203 in the overlapping area 301 are determined to be ' _hoffset ' and 'h', respectively.

6 is a graph illustrating the Gaussian curve for an object in accordance with embodiments of the present invention.

Referring to FIG. 6, the Gaussian curve 601 may include a Gaussian distribution. Here, the Gaussian distribution is a continuous probability distribution that can be used as a first approximation to describe real-valued random variables that tend to converge to a single average value. And the Gaussian curve 601 allows for convenient selection for modeling a wide variety of random variables. The alpha mask generation may be induced by a Gaussian profile, and the Gaussian curve may be determined by the following equation.

Here, 'a' represents the peak / maximum value of the curve, 'b' represents the average value of the curve, for example, the center position of the peak, and 'c' represents the width of the Gaussian curve .

Here, HWHM denotes a Half Width at Half Maximum, and 'e' denotes an Euler's number.

Information about the object 203 in the overlap area 301 can be provided to the image processing apparatus 102 in the form of width, height, and coordinates. And the values of the variables a, b, and c for the object 203 in the overlap area 301 can be determined by the following equations.

Here, 'h' represents the height of the object 203 in the overlap area 301, 'x _offset ' represents the width from the beginning of the overlap area 301 to the object 203, 'w _object 'Indicates the width of the object 203 existing in the overlap area 301.

 For example, some background information behind the object 203 may be generated by considering twice the width of the object 203 as the HWHM. The value of HWHM is determined by the user or the external system.

7 is a graph showing a slope of a simple linear slope according to embodiments of the present invention.

이고, a_i, b_i, ... , c_i는 앞에서 언급된 기법에 의해 i번째 물체에 대해 계산될 수 있다.7, the overlap area 301 may include a plurality of objects, such as object 1 701, object 2 702, and object 3 703. The Gaussian curve can be determined for each of the objects 701, 702, and 703. For example, the last Gaussian curve can be determined by superposition of each Gaussian curve about each object 701. The last Gaussian curve can be given as y = y1 + y2 + y3 + ... + yn. here,

, And a _i , b _i , ..., c _i can be calculated for the i th object by the above-mentioned technique.

As another example, the last Gaussian curve for a plurality of objects present in the overlap area 301 may be determined as the minimum value of the overlapping curves. That is, the last Gauss curve can be determined as y = min (y1, y2, y3, ... yn).

8 is a diagram illustrating a plurality of objects in an overlap region according to embodiments of the present invention.

8, the last guiding curves for the plurality of objects 801, 802, 803 in the overlap area 301 can be determined by summing the individual Gaussian curves values to obtain a basic profile.

Once the basic profile is determined, the values in the alpha mask can be determined. Where the values of the alpha mask determine the transparency information for each pixel. The alpha mask specifies the transparency of the pixels across the overlap region 301. All alpha values on the Gaussian curve and below the Gaussian curve may be considered as '1' to make image 2 opaque so that image 1 on the object of interest 203 is not visible. Here, the opaque image 2 can protect and emphasize the object 203 of interest.

The remaining alpha mask values can be varied from 0 to 1 so that the pixel values gradually increase from ' 0 ' until ' 1 ' until the pixel values reach the Gaussian curve.

9 is a diagram illustrating a composite guide profile constructed from individual curves of objects in accordance with embodiments of the present invention.

와 같이 주어진다. 여기서, 'row=0'은 중첩 영역(301)의 정상에 있는 경우이며, 아래쪽으로 증가한다. 'd'는 'row=0'에서 가우스 곡선까지의 거리이며,

는 알파의 최대 값을 나타낸다. 일반적으로, 이 값은 '1'이지만, 구현 시 부동 소수점 산술을 피하기 위해 '255'와 같은 더 큰 수를 취할 수 있다.Referring to FIG. 9, the slope 901 for the alpha value can be varied to obtain various results. The alpha value for the linear slope is

As shown in Fig. Here, 'row = 0' is at the top of the overlap region 301 and increases to the bottom. 'd' is the distance from 'row = 0' to the Gaussian curve,

Represents the maximum value of alpha. Generally, this value is '1', but implementations can take a larger number such as '255' to avoid floating-point arithmetic.

와 같이 주어질 수 있다. 여기서, k는 양의 실수이며, row ∈ [0, d]인 조건 하에

이다.For example, a non-linear slope can be considered to avoid the ghosting effect on the synthesized image. The alpha value for the nonlinear gradient is

As shown in FIG. Where k is a positive real number, and under the condition row ∈ [0, d]

to be.

Here, if k = 1 is assigned to the equation for the nonlinear gradient, it can be a formula for the linear gradient. And the equation for the nonlinear gradient has a concave shape with respect to the y-axis. The curve gradually increases from the alpha value of '0' to the maximum concave point and begins to increase sharply until the alpha value is '1'. This concave curve can reduce the ghosting effect by eliminating mid-range alpha values.

로 계산될 수 있다.After determining the alpha mask, images may be synthesized using synthetic techniques such as alpha blending, Poisson blending, and the like. For example, in the alpha blending technique, the last pixel of the overlap region 301

Lt; / RTI >

As another example, the object 203 of image 2 existing in the overlapping area 301 can cover almost the entire image 1 so that most of the overlap area 301 can be occupied by the object 203 or a part of the object 203. [ Can be covered. And region 2 may have relatively small data, which is in stark contrast to overlapping region 301. The alpha mask profile is used to obtain the final composite image.

10 is a diagram showing an image synthesized by alpha mask generation according to embodiments of the present invention.

Referring to FIG. 10, a first image 1001 and a second image 1002 represent input images. When the first image 1001 is combined with the second image 1002, the overlapping area can be determined. And the overlap region may include a plurality of objects of the second image 1002. [ A plurality of objects of the second image 1002 need to be protected when they are combined with the first image 1001 in the overlapping area. After performing the above-described synthesis method with alpha mask generation, a plurality of objects of the second image 1002 can be protected. And emphasize a plurality of objects of the second image 1002 to emphasize visual effects like the composite image 1003. [

Embodiments of the present invention may be implemented through at least one software program running on at least one hardware device and performing network management functions to control network elements. The network elements shown in Figure 1 include blocks that can be at least one of a hardware device or a combination of a hardware device and a software module.

Embodiments of the present invention provide a method for enhancing a user experience on a computing device by enabling customization of the application by allowing one or more resident client entities to negotiate with one or more client executing entities or servers in terms of customizable applications And systems. Therefore, it is understood that the scope of the present invention extends to computer-readable means including such programs and messages, and when the program is run on a server, mobile device, or appropriate programmable device, such computer- And program code means for performing the above steps. The method may be implemented, for example, through a software program written in a high speed integrated circuit hardware description language (VHDL) or in a preferred embodiment, Software modules, or one or more VHDLs. The hardware device may be a programmable portable device. The device may also include, for example, a hardware means such as an ASIC, a combination of hardware and software such as an ASIC and an FPGA, or a means that may be at least one memory and at least one microprocessor including software modules. Embodiments of the method described herein can be implemented in hardware and software in part. Alternatively, the present invention may be implemented on different hardware devices using a plurality of CPUs, for example.

It will be apparent to those skilled in the art that various changes and modifications may be made without departing from the spirit and scope of the present invention as defined by the appended claims. will be. It will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined by the appended claims.

Claims (12)

A method of compositing an image in an image processing apparatus,
Determining whether at least one object included in the images exists in an overlapping area between the plurality of images;
Generating a Gaussian curve for the object if the object exists;
Determining an alpha mask for the overlap region;
And synthesizing the images based on the alpha mask.
The method according to claim 1,
Wherein the Gaussian curve is generated using the following equation.
&Quot; (1) "

Wherein a represents a maximum value of the Gaussian curve, b represents an average value of the Gaussian curve, c represents a width of the Gaussian curve,

, The HWHM represents a Half Width at Half Maximum, and the e represents an Euler number.
The method according to claim 1,
Wherein the variables for the object are determined by the following equation.
&Quot; (2) "

H is the height of the object in the overlap region, x _offset is the width from the beginning of the overlap region to the object, w is the width of the _object in the overlap region, Indicate.
The method according to claim 1,
Wherein the process of generating the Gaussian curve is a process of generating the Gaussian curve by superimposing Gaussian curves for the at least two objects when the object has at least two objects.
The method according to claim 1,
Wherein the generating of the Gaussian curve is a process of generating the Gaussian curve using minimum values of Gaussian curves for the at least two objects when the object has at least two objects.
The method according to claim 1,
Wherein the alpha mask is determined by the following equation.
&Quot; (3) "

Wherein k is a positive real number, d represents a distance from row = 0 to the Gaussian curve,

Represents the maximum value of the alpha mask. An apparatus for synthesizing an image in an image processing apparatus,
A display section,
The method comprising: confirming whether there is at least one object included in the images in an overlapping area between a plurality of images, and if the object exists, generating a Gaussian curve for the object; And an image synthesizer for determining an alpha mask and synthesizing the images based on the alpha mask.
8. The method of claim 7,
Wherein the Gaussian curve is generated using the following equation.
&Quot; (4) "

Wherein a represents a maximum value of the Gaussian curve, b represents an average value of the Gaussian curve, c represents a width of the Gaussian curve,

, The HWHM represents a Half Width at Half Maximum, and the e represents an Euler number.
8. The method of claim 7,
Wherein variables for the object are determined by the following equations.
Equation (5)

H is the height of the object in the overlap region, x _offset is the width from the beginning of the overlap region to the object, w is the width of the _object in the overlap region, Indicate.
8. The method of claim 7,
Wherein the image synthesizing apparatus generates the Gaussian curve by superimposing Gaussian curves for the at least two objects when the object has at least two objects.
8. The method of claim 7,
Wherein the image synthesizing apparatus generates the Gaussian curve using minimum values of Gaussian curves for the at least two objects when the object has at least two objects.
8. The method of claim 7,
Wherein the alpha mask is determined by the following equation.
&Quot; (6) "

Wherein k is a positive real number, d represents a distance from row = 0 to the Gaussian curve,

Represents the maximum value of the alpha mask.

Images