CN100386772C - Image blending by guided interpolation - Google Patents

Image blending by guided interpolation Download PDF

Info

Publication number
CN100386772C
CN100386772C CN 200480003010 CN200480003010A CN100386772C CN 100386772 C CN100386772 C CN 100386772C CN 200480003010 CN200480003010 CN 200480003010 CN 200480003010 A CN200480003010 A CN 200480003010A CN 100386772 C CN100386772 C CN 100386772C
Authority
CN
China
Prior art keywords
image
target
source
boundary
interpolation
Prior art date
Application number
CN 200480003010
Other languages
Chinese (zh)
Other versions
CN1745386A (en
Inventor
A·布雷克
M·冈根特
P·佩雷斯
Original Assignee
微软公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Priority to US45007803P priority Critical
Priority to US60/450,078 priority
Priority to US10/417,937 priority
Application filed by 微软公司 filed Critical 微软公司
Publication of CN1745386A publication Critical patent/CN1745386A/en
Application granted granted Critical
Publication of CN100386772C publication Critical patent/CN100386772C/en

Links

Abstract

利用受指导的的内插计算混和的结果图像,以改变目标域内的图像数据。 By using guided interpolation calculation blended result image, to change the image data of the target domain. 可基于依赖于或不依赖于源图像的受指导的内插来改变目标域。 Interpolation may be changed based on the target domain guided by the dependent or dependent on the source image. 在将源区域混和到图像的目标区域时,受指导的的内插消除或最小化插入的区域和图像其余部分间的明显接缝。 When the source area to the target area image mixed, guided by the interpolation eliminates or minimizes apparent and the image insertion area of ​​the joints between the remaining portion. 可使用各种内插指导,以在混和的结果图像中产生不同的效果。 Guidance may be provided with a variety of interpolation, to produce different effects in the blended result image. 此类内插指导可包括,但不限于,恒等指导、平滑指导、过滤指导、透明指导、掩模指导、以及各种指导的加权平均组合。 Such guidance may include the interpolation, but are not limited to, the identity guide, a smooth guide, a filter guide, a transparent guide, a mask guide, and a weighted average combination of various guidance.

Description

通过受指导内插的图像混和 By blending the interpolated by the image guidance

相关申请 RELATED APPLICATIONS

此申请要求美国临时专利申请第60/450,078号,题为"Image Blending By Guided Interpolation"(受指导的内插的图像混和)的优先权,该申请通过整体引用其所公开并教导的内容而特别地包括在此。 This application claims the benefit of US Provisional Patent Application No. 60 / 450,078, entitled "Image Blending By Guided Interpolation" (under the direction of the interpolated image blending) priority, which is incorporated herein by reference its disclosure and taught by specially overall content included in this.

技术领域 FIELD

本发明一般涉及图像编辑,尤其涉及混和图像。 The present invention relates generally to image editing, and more particularly a blended image. 发明背景 BACKGROUND OF THE INVENTION

编辑数字图像,尤其是数字照片,是家庭和专业环境中常见的活动。 Editing digital images, especially digital photos, home and professional environments common activities. 一种常见的任务包括将一块图像贴到另一图像中,或者贴到同一图像的其它位置。 One common task includes an image of another image attached to, or attached to other locations of the same image. 此操作用于修复、修正和修改数字图像。 This operation is used to repair, revise and modify digital images. 例如,在图1中,用户希望将来自源图像102的太阳100插入到目标图像104 中。 For example, in FIG. 1, the user wishes to sun image 102 from the source 100 into the target image 104. 在简单的复制粘贴操作中,用户将在源图像102中沿着太阳100的圆周进行选择或裁剪、复制所裁剪的区域、并将所裁剪的区域贴到目标图像104内的期望位置。 In a simple copy and paste operation, the user will be in the source image 102 along a circumference of the sun or the crop 100 is selected, the copy area cropped, and the cropped region attached to a desired position within the target image 104. 或者,可以用类似方式使用简单的拖放操作。 Alternatively, in a similar manner using a simple drag and drop operation. 然而,在这两种方法中,插入的所裁剪区域和目标图像104的剩余部分之间的边界(或称"接缝")会显得很不连贯并反映裁剪操作中的任何不规则。 However, in both methods, the remaining portion of the boundary between the insert and the cropped area of ​​the target image 104 (or "joints") will become very consistent and reflect any irregularities in the cutting operation. 此外,当插入到由灰蒙蒙的背景表示天空的目标图像104时,在裁剪操作中所捕捉的源图像102中太阳的背景的任何部分(即,彩色图像中纯浅蓝色)将表现为生硬的不连续。 In addition, when inserted into the sky is represented by a gray background object image 104, in the cutting operation in the captured image 102 in any portion of the source of the sun's background (i.e., pure blue color images) will appear to be stiff discontinuous. 换言之,所插入的裁剪部分不会混和到目标图像104中。 In other words, the inserted clipping would not blend into the destination image portion 104.

一种称作"克隆"的替换方法涉及在源图像内定义的定位点。 A technique called "cloning" of the alternative method involves locating the point source image within the definition. 在目标图像上使用画笔控制来定义目标图像中应将源图像数据复制到其中的区域。 To define the target image, the image data should be copied to the source region of the brush is used to control where on the target image. 结果类似于复制粘贴或拖放替换方法。 Results similar copy-paste or drag and drop an alternate method.

此外,其它图像编辑操作也可涉及改变图像中选中区域的外观。 In addition, other image editing operations can also involve altering the appearance of the image in the selected area. 例如,可以 For example,

平化图像某区域的纹理、可以使背景或前景脱色、或者可以在图像的某区域上进行照明变化。 The texture of an image of an area level, bleaching can background or foreground, or may be an area on the illumination variation image. 然而,此类编辑操作也会引入所编辑区域和原始图像其余部分间接缝的问题。 However, such editing operations also introduce problems editing area and the rest of the original image of joints.

为了改进这些现有方法的结果,可以进行局部"羽化"操作来隐藏所插入的裁剪部分和目标图像背景间的接缝。 In order to improve the results of these existing methods may be local "feathering" operation to hide the seams between the inserted clipping and the destination image background portions. 这样的羽化方法通常涉及在接缝处对图像数据进行模糊或涂抹,以逼近所插入的裁剪部分和目标图像背景间的平化过渡。 Such methods generally involve the emergence of blurring or smearing the image data at the seam to approximate the level of the transition between the inserted clipping and the destination image background portions. 然而, 羽化操作可能产生不能令人满意的结果,因为接缝只不过是被模糊的图像数据取代。 However, feathering operation may produce unsatisfactory results because the seam is only substituted blurred image data. 此外,现有方法不能提供各种混和模式以允许用户将所插入的图像部分最好地集成到目标图像中。 Moreover, various conventional methods can not provide a mix mode to allow the user to insert the image portion is preferably integrated into the target image.

因此,在许多普通情形中,现有的裁剪和克隆方法不能令人满意地提供灵活和无缝的图像插入与编辑。 Thus, in many common cases, the existing clipping and cloning methods are not satisfactorily provide flexible and seamless image insertion and editing.

发明概述 SUMMARY OF THE INVENTION

本发明的各实施例通过使用受指导的内插改变目标域内的图像数据来提供混和的结果图像,从而解决了所讨论的问题。 EXAMPLE result image is provided by using a mixed interpolation of the image data is changed by the guide of the target region of the respective embodiments of the present invention to solve the problems discussed. 可以基于依赖或独立于源图像的受指导的内插器来改变目标域。 May vary based on the target domain interpolator dependent or independent of the source of the image-receiving guidance. 可以使用各种内插指导在混和的结果图像中产生不同的效果。 Insertion guide can produce different effects in the blended result image within the variety used.

在本发明的实现中,提供了制造物品作为计算机程序产品。 In the implementation of the invention, there is provided an article of manufacture as a computer program product. 计算机程序产品的一实施例提供了计算机程序存储介质,它是计算机系统可读的,并编码了从目标图像计算混和的结果图像的计算机程序。 An embodiment of a computer program product provides a computer program storage medium which is readable by a computer system and encoding a computer program calculates blended result image from a target image. 可以用计算机数据信号的形式提供计算机程序产品的另一实施例,该计算机数据信号由计算系统在载波中具体化,并编码了从目标图像计算混和的结果图像的计算机程序。 Another computer program product may be provided in the form of a computer data signal embodiment, the computing system by a computer data signal embodied in a carrier wave, a computer program and results of encoding the image calculated from the target image blending.

该计算机程序产品编码了在计算机系统上执行、用于从目标图像计算混和的结果图像的计算机过程的计算机程序。 The computer program product code execution on a computer system, a computer program for a computer process blended result image calculated from the target image. 在目标图像中定义了具有边界的目标域。 It defines the target domain having a boundary in the target image. 提供了包括内插指导和与目标域的边界相关联的边界条件的受指导的内插器。 It provides guidance and including the interpolation target domain boundary conditions associated with the boundary under the guidance of the interposer. 基于受指导的内插器计算混和的结果图像,以满足目标域边界处的边界条件,并将内插指导和混和的结果图像的梯度间跨越目标域的差值最小化。 The difference between the calculated gradient across the target domain based mixing guided by the result of the image interpolator to satisfy the boundary condition at the boundary of the destination domain, and the insertion guide and mixed in the resultant image is minimized.

在本发明的另一实现中,提供了一种从目标图像计算混和的结果图像的方法。 In another implementation of the present invention, there is provided an image computed results from the target image blending method. 在目标图像中定义具有边界的目标域。 Is defined in the target image has a target domain boundaries. 提供了包括内插指导和与目标域的边界相关 It provided, including interpolation and guidance related to the target domain boundary

联的边界条件的受指导的内插器。 Guided by the boundary conditions associated with the interposer. 基于受指导的内插器计算混和的结果图像,以满足目标域边界处的边界条件,并将内插指导和混和的结果图像的梯度间跨越目标域的差值最小化。 The difference between the calculated gradient across the target domain based mixing guided by the result of the image interpolator to satisfy the boundary condition at the boundary of the destination domain, and the insertion guide and mixed in the resultant image is minimized.

在本发明又一实施例中,提供了一种用于从目标图像计算混和的结果图像的系统。 In yet another embodiment of the present invention, there is provided a system for calculating blended result image from the target image. 域定义模块在目标图像中定义具有边界的目标域。 Domain definition module defines the target domain having a boundary in the target image. 内插指导数据存储提供了包括内插指导和与目标域的边界相关联的边界条件的受指导的内插器。 The interpolation guide datastore provides guidance including the interpolation target domain and a boundary condition associated with the boundary guided by the interpolator. 混和计算模块基于受指导的内插器来计算混和的结果图像,以满足目标域边界处的边界条件, 并将内插指导和混和的结果图像的梯度间跨越目标域的差值最小化。 A difference across the destination domain between the gradient minimized mixing calculation module calculates based mixed guided by the result of the image interpolator to satisfy the boundary condition at the boundary of the destination domain, and the interpolation guide and a blended result image.

在阅读以下详细描述并查看相关附图后,表现本发明特性的这些及各种其它特征,连同其它优势,将会变得显而易见。 After the description of the accompanying drawings and detailed view upon reading the following, the performance characteristics of the present invention These and various other features, along with other advantages, will become apparent.

附图简述 BRIEF DESCRIPTION

图1描绘了本发明的一个实施例中源图像、目标图像、以及混和的结果图像的示意性表示。 FIG 1 depicts one embodiment of the present embodiment of the invention the source image, the target image, and a schematic representation blended result image.

图2描绘了本发明的一个实施例中两个一维函数的概念上的混和。 2 depicts a conceptual blending one embodiment, two one-dimensional functions of the embodiment of the present invention. 图3示出了本发明的一个实施例中两个二维图像区域的混和。 FIG. 3 shows the invention in a mixture of two-dimensional image regions embodiment of FIG. 图4描绘了本发明的一个实施例中的源图像、目标图像、和结果图像。 4 depicts a source image in this embodiment, the target image, the image and the results of the present invention. 图5描绘了本发明的一个实施例中的混和系统。 FIG 5 depicts an embodiment of the present invention the hybrid system. 图6描绘了本发明的一个实施例中用于混和的操作。 FIG 6 depicts an embodiment for the mixing operation of the embodiment of the present invention. 图7描绘了本发明的一个实施例中源自透明内插指导的混和结果。 7 depicts one embodiment of the present invention is derived from the result of mixing the transparent interpolation guide. 图8描绘了本发明的一个实施例中源自掩模内插指导的混和结果。 8 depicts an embodiment of the present invention, the blended result in the insertion guide from a mask embodiment. 图9示出了本发明的一个实施例中混和到目标图像中。 FIG 9 shows an embodiment of the present invention is mixed into the target image embodiment. 图IO示出了用于实施本发明的一个实施例的示例性系统。 FIG IO illustrates an exemplary embodiment of a system for one embodiment of the present invention.

较佳实施例的详述 Detailed Description of the preferred embodiment

本发明的各实施例使用受指导的内插改变目标中图像数据,从而提供了混和的结果图像。 Embodiments of the present invention uses guided by the interpolation target image data changes, thus providing a blended result image. 可以基于依赖或独立于源图像的受指导的内插器来改变目标域。 May vary based on the target domain interpolator dependent or independent of the source of the image-receiving guidance. 当将源区域混和到图像的目标域中时,受指导的内插能够消除或最小化所插入区域和图像的其余部分间的明显接缝。 When the source area to the target domain mixed image, guided by interpolation can be eliminated or significantly joints between the rest region and the image of minimizing inserted. 此外,可以使用各种内插指导以在 In addition, you can use within a variety of plug guidance to the

混和的结果图像中产生不同效果。 The blended result image have different effects. 此类内插指导可包括,但不限于,恒等指导、 平滑指导、过滤指导、透明指导、掩模指导、以及各种指导的加权平均组合。 Such guidance may include the interpolation, but are not limited to, the identity guide, a smooth guide, a filter guide, a transparent guide, a mask guide, and a weighted average combination of various guidance. 图l描绘了本发明的一个实施例中的源图像、目标图像和混和的结果图像。 Figure l depicts a source image in an embodiment of the present invention, the result of the target image and blended.

源图像102包括含有太阳的源区域100。 The image source 102 includes a source region 100 containing a sun. 用户希望将源区域100插入并混和到目标图像104中,后者最初在位置108处不包括所插入的太阳。 The user wishes to insert the source region 100 and mixed to the target image 104, which does not initially include the inserted sun at location 108. 在第一操作中, 用户通过诸如在源区域100周围拖曳出轮廓,来定义源区域100。 In the first operation, the user such as by dragging the region around the source profile 100, 100 to define a source region. 可以使用定义源区域100的替换方法,包括但不限于,为克隆画笔设置定位点、在图像中选择元素、指定源区域的色彩特性、或在源图像102中指定坐标等。 Alternative methods can be used to define the source region 100, including but not limited to, a positioning point cloning brush, selecting an element in the image, the color attribute specifies the source region, or specifying coordinates in the source image 102 and the like.

一般而言,数字图像由像素值表示。 Generally, a digital image represented by a pixel value. 例如,黑白的数字图像可由像素亮度值矢量来表示,其中每个矢量元素对应于单个像素。 For example, a black and white digital image may be represented by the pixel luminance value vector, wherein each vector element corresponds to a single pixel. 相反,彩色图像的数字图像数据可以用每个像素的多个值来表示。 In contrast, the image data of the color digital image may be a plurality of values ​​of each pixel is represented. 例如,在RGB (红、绿、蓝)色彩空间中,每个像素由3个色彩通道中的每一个的亮度值来呈现。 For example, in the RGB (red, green, blue) color space, each pixel is presented by the luminance values ​​of the three color channels each. 还构想了其它色彩空间,包括但不限于LC1C2和CIE-lab空间。 Other color spaces are also contemplated, including but not limited to LC1C2, and CIE-lab space.

应当理解,仅将源区域IOO复制粘贴到目标图像104中通常会导致围绕太阳的不规则蓝天色的边界或接缝110,它覆盖目标图像104的较暗的夜空背景112。 It should be understood that only the copy source area to the target image 104 pasted IOO typically leads to an irregular blue color around the sun of the boundary or joint 110, which covers the darker night sky background 104 of the target image 112. 与目标图像相比,接缝通常表现为色彩、亮度、特征、纹理等等中的突变。 Compared to the target image, the seam usually presents color, brightness, features, texture, etc. mutations. 如此,要产生较逼真的结果图像,所插入源区域104的混和可以模糊或消除结果图像106中的接缝110的可见性。 Thus, to produce a more realistic result image, a source region 104 may be inserted blur or a mixture eliminate visible seams 110 of the image 106 results. 因此,如结果图像106中所示,源区域IOO被插入并混和到目标图像104中,以产生逼真的效果。 Thus, as shown in the result image 106, the source region and mixed IOO is inserted into the target image 104 to produce a realistic effect.

图2描绘了本发明的一个实施例中两个一维函数的概念上的混和。 2 depicts a conceptual blending one embodiment, two one-dimensional functions of the embodiment of the present invention. 关于图2 的该描述提供了涉及一维"图像"的混和操作的概念性演示。 With respect to the description of FIG. 2 provides a conceptual demonstration relates to a one-dimensional "image" blending operation. 但是,应当理解,在本发明的范畴内构想了各种混和操作,包括但不限于,相对于本文中其它附图所描述的二维混和操作。 However, it should be understood that the various mixing operations contemplated within the scope of the invention, including but not limited to, with respect to the two-dimensional blending operations described in other figures herein.

在目标曲线图200中示出了表示一维目标"图像"的第一亮度函数A。 In the target graph 200 illustrates a representation of a target-dimensional "image" of the first luminance function A. 在源曲线图202中示出了表示要插入到曲线图200的目标图像中的一维源"图像"的第二亮度函数B。 In the source graph 202 shows a representation of the source to be inserted into a one-dimensional image of the target graph 200 in an "image" of the second intensity function B. 结果曲线图204中示出了表示一维混和结果"图像"的混和的结果函数。 The results are shown in the graph 204 represents the result of a one-dimensional blended result function "image" blended.

虚线206表示源区域和目标区域的边界。 Dashed line 206 represents a boundary of the source and destination zones. 在本发明的另一个实施例中,源区域和目标域可以具有完全不同的尺寸和形状(例如,矩形、圆形、不规则形),两 In another embodiment of the present invention, the source region and the target domain may have different sizes and shapes (e.g., rectangular, circular, irregular), two

者都可由系统改变,从而对于混和更为相容。 Who can be changed by the system, and thus more compatible for mixing.

在本发明的一个实施例中,用户希望使用混和操作将源区域合并到目标域中, 从而目标域边界处的图像数据被混和以表现逼真。 In one embodiment of the present invention, the user wishes to use the combined operation of mixing the source region to the target domain so that the image data at the boundary of the destination domain are blended to represent realistic. 在所示出的实施例中,这通过把第一亮度函数A的源区域连到第二亮度函数B的边界点来概念性地示出。 In the illustrated embodiment, this is shown conceptually by the source region of the first intensity function A is connected to the second boundary point B of the brightness function. 如此,混 So, mixed

和的结果函数示出了两个亮度函数的组合。 The resulting function and shows a combination of the two intensity function. 为了消除、最小化、或减少接缝的出现, In order to eliminate, minimize, or reduce the appearance of seams,

各组成函数的边界条件被混和,从而当各函数逼近边界点208和210时,每个组成函数的值相等。 Boundary conditions for each composition was mixed function, such that when the respective function approximation boundary points 208 and 210, each value of the function equal to the composition.

图3示出了在本发明的一个实施例中两个二维图像区域的混和。 FIG. 3 shows a two-dimensional image mixing region embodiment of the present invention in one embodiment. 为了扩展图2 的一维混和的概念性描述,图3中提供了二维混和的示例。 To extend the conceptual description of the one-dimensional blending of FIG. 2, FIG. 3 provides an example of two-dimensional mixing. 目标图像300示出黑暗的水面。 Target image 300 shows a dark water surface. 二维亮度曲线图302中示出了与目标图像300相关联的对应亮度数据,其中亮度可包括单色亮度(诸如黑白图像中的白色,或者彩色图像中的单个色彩通道),或者某种其它外观信号(例如,LC1C2的分量或聚合值)。 Two-dimensional luminance graph 302 shows the corresponding luminance data and target image 300 is associated, which may include a monochromatic luminance brightness (white, such as black and white images, or a single color channel in a color image), or some other appearance signal (e.g., LC1C2 value or polymeric component). 例如,注意当目标图像300在接近顶部处变得较浅时,亮度曲线图302中所示的白色的亮度增加了。 For example, note that when the target image 300 becomes lighter near the top, shown in white luminance in the luminance graph 302 increases.

在中间图像304中,包含在不同水面上太阳的倒影的源区域312被插入到该图像中。 Intermediate image 304 includes a source region 312 of the reflection of the sun is inserted into the image on a different surface. 在所示的实施例中,将目标区域定义为被源区域312重叠的图像的区域。 In the illustrated embodiment, the target region is defined as the area of ​​the image 312 overlaps the source area. 在混和之前,至少部分因为水的颜色的差异,源区域和目标图像的其余部分间的接缝非常明显,如中间图像304中所示。 Before mixing, at least in part because of differences in the color of the water, the joints between the source region and the rest of the target image is very clear, as shown in the intermediate image 304. 这在非混和的亮度曲线图306中也可看出来, 其中涉及源区域312的亮度数据在314处是不连贯地可见的。 This can also be seen in the non-blended luminance graph 306, wherein the luminance data 312 is directed to the source region 314 is not consistently visible.

在混和的结果图像308中,源区域被混和到目标区域中,以提供太阳在水面上倒影的逼真图像。 In the blended result image 308, the source region to a destination region is blended to provide a realistic image of the reflection of the sun on the water. 在混和的结果亮度曲线图310中明显可见源区域中图像数据亮度的变化,其中所插入的源区域的亮度数据被混和,以在此第一区域的边界处匹配目标图像的亮度数据,从而降低了源区域的总"亮度"。 In the result, the luminance graph of the mixing region 310 visible source image data of a change in luminance, wherein the luminance data of inserted source region is mixed, at the boundary of this first region of the target image matches the brightness data to reduce overall "brightness" of the source region.

图4描绘了本发明的一个实施例中的源图像、目标图像和混和的结果图像。 4 depicts an embodiment of the present invention, the embodiment of the source image, the result of the target image and blended. 目标图像400 (以彩色图像)示出海滩处浅绿的海面。 400 target image (color image) shown at a light green sea beaches. 源图像402 (以彩色图像) 示出水显得浅蓝的游泳池中的两个孩子。 Source image 402 (color image) shows the water appears light blue of the swimming pool two children. 出于此讨论的目的,并为了将图像特征映射到本文中所描述的算法,目标图像400被称作,目标图像400中的目标区域404被称作"Q"。 For purposes of this discussion and to map image features to the algorithms described herein, is referred to as a target image 400, target area 404 is referred to as a target image 400 "Q". 目标区域404的边界412被称作""。 Boundary of the target area 404 is referred to as 412 "." 类似地,源图像402被称作"g*",源区域416的图像数据被称作g。 Similarly, the source image 402 is referred to as "g *", the image data source region 416 is referred to as g. 在混和的结果图像410 中,源区域416被混和到目标区域404中,以产生包含图像区域418和420 (Q内 In blended result image 410, the source region 416 is blended into the target area 404, to produce an image containing regions 418 and 420 (Q within

部的混和图像数据/,和/*的其余部分,即目标图像在目标区域外的部分)的合成结果图像。 Mixing the image data /, and / * the remaining portion, i.e., part of the target image outside the target area) of the portion of the resulting composite image. 注意:混和的结果图像是通过按顺序将包含每个孩子的源区域插入到目标图像中来生成的。 Note: The blended result image is obtained by sequentially comprising the source region of each child is inserted into the target image is generated.

在本发明的一个实施例中,混和操作是使用以下函数最小化作为受指导的内插器(或"最小化器")来实现的: In one embodiment of the present invention, the mixing operation is minimized as guided by the interposer (or "minimizer") to achieve the following functions:

它具有边界条件/l^^/t,其中 It has boundary conditions / l ^^ / t, wherein

<formula>formula see original document page 10</formula> <Formula> formula see original document page 10 </ formula>

是梯度算子。 Is the gradient operator. 此外,可以使用其它内插器,诸如最大化函数和最优化成给定目标、阈值或范围的函数。 In addition, other interpolators, such as to maximize the function and optimized to a given target, or a function of the threshold range.

相关联的欧拉一拉格朗日(Euler-Lagrange)方程,即受指导的内插器所满足的条件,被认为是具有狄利克雷(Dirichlet)边界条件的泊松(Poisson) Euler-Lagrange (Euler-Lagrange) associated equation, i.e., by the condition of the advice interpolator satisfied, Poisson is considered (the Poisson) with Dirichlet (the Dirichlet) boundary conditions

方程: equation:

Q上的A/,Ag'其/L^/t, A on the Q /, Ag 'which / L ^ / t,

■^^ a'二」是拉普拉斯(Laplacian)算子。 ■ ^^ a 'two' is a Laplace (Laplacian) operator. 受指导的内插方法可由以下函数最小化更概括地表示为受指导的内插器: Interpolation method may be guided by the following function is minimized More generally designated by guided interpolator:

<formula>formula see original document page 10</formula> <Formula> formula see original document page 10 </ formula>

它具有边界条件/l^^/t其中v表示内插指导,/表示"内插式"。 It has boundary conditions / l ^^ / t where v represents an interpolation guide, / indicates "the plug-in." 积分示意 Points schematic

该函数最小化是以下泊松方程的唯一解: Q上的Ay-divv,其/"=, The function minimization is the unique solution of Poisson's equation the following: Ay-divv on Q, which is / "=

其中 among them

<formula>formula see original document page 10</formula> <Formula> formula see original document page 10 </ formula>

其中 among them

<formula>formula see original document page 10</formula> <Formula> formula see original document page 10 </ formula>

是v-[w,v]的散度, Is v- [w, v] of divergence,

v的各种选择允许以各种方法进行混和,如以下示例内插指导表中所示, v Various options allow mixing, the following example is inserted in the table shown guidance in various ways,

尽管在本发明的范畴内还构想了其它内插指导。 Although also contemplated within the scope of the invention other interpolation guide.

<table>table see original document page 10</column></row> <table> <Table> table see original document page 10 </ column> </ row> <table>

<table>table see original document page 11</column></row> <table>例如,恒等内插指导/导致目标图像内无变化(因此,/=/*)。 <Table> table see original document page 11 </ column> </ row> <table> For example, the identity interpolation guide / cause no change in the target image (Thus, / = / *). 替换地, 平滑内插指导s改变目标图像。 Alternatively, the smooth interpolation guide target image s changed. 与平滑内插指导s相关联的欧拉一拉格朗日方程是:<formula>formula see original document page 11</formula>并且被认为是具有狄利克雷边界条件的拉普拉斯方程。 Smooth interpolation guide and the s associated Euler-Lagrange equation is: <formula> formula see original document page 11 </ formula> and is considered to Laplace's equation with Dirichlet boundary conditions. 具有狄利克雷边界条件 With a Dirichlet boundary condition

的拉普拉斯方程的唯一解常被称作"膜内插式(membrane interpolant)"。 The only solution to Laplace's equation is often referred to as "plug-in film (membrane interpolant)". 在每个色彩通道中解决此问题的离散化形式导致域Q中非常平滑或模糊的图像区域。 To solve this problem in each color channel in the form of discrete domains in Q results in a very smooth or blurred image region. 平滑内插指导5将该域内部的梯度最小化("空指导"),并导致位于域边界上的色彩的非常平滑的内插。 The smooth interpolation guide gradient within the portion 5 is minimized ( "air guide"), and lead positioned on the color gamut boundary is smooth interpolation. 平滑内插指导^可由计算机图形区域填充算法使用,包括多边形或任意形状的填充算法。 Smooth interpolation guide ^ by the computer graphics area filling algorithm, the algorithm comprising filling a polygon or arbitrary shape.

透明和掩模内插指导的示例性效果分别相对于图7和8示出并讨论。 And the mask transparent interpolation guide exemplary effect with respect to FIGS. 7 and 8 shown and discussed. 图5描绘了本发明的一个实施例中的混和系统。 FIG 5 depicts an embodiment of the present invention the hybrid system. 图像编辑系统500接收诸如目标图像的图像数据等的目标504。 The image editing system 500 receives a target 504 such as image data of the target image and the like. 目标选择模块508接收对目标504内的目标区域的选择。 Object selection module 508 receives a selection of a target region within the target 504. 如前所述,目标区域可通过粘贴操作、克隆操作、画轮廓操作、或者选择目标图像的区域的其它方法来定义。 As described above, the target region may be by gluing operation, cloning manipulations contouring operation, or other methods of selection target image region is defined.

在至少一个实施例中,图像编辑系统500还接收诸如源图像等源502。 In at least one embodiment, the image editing system 500 also receives source 502, such as source images. 源选择模块506接收对源504内的源区域的选择。 Source selection module 506 receives a selection of a source region within the source 504. 但是,应当理解,诸如那些涉 However, it should be understood that, such as those involving

及平滑内插指导或掩模内插指导的实施例等替换实施例不需要源或源选择模块,因为该内插指导独立于任何源。 Example source and does not need other alternative embodiments the smooth interpolation guide or a masking interpolation guide in the embodiment or a source selection module because the interpolation guide is independent of any source within.

域定义模块510接收一个或多个选择,并定义目标域(可能还有源域)。 Domain definition module 510 receives one or more selection, and define the destination domain (and possibly the source domain). 在一个实施例中,域定义可以是图像上X和Y坐标的范围的形式,尽管在本发明的范畴内构想了定义域的其它方法,包括但不限于图像位置的数组以及区域定义算法。 In one embodiment, the domain definition may be in the form of a range of the X and Y coordinates of the image, although within the scope of the invention contemplates other methods of domain, including but not limited array and a region defined algorithm to the image position.

图像数据收集器511收集输入图像中在本文中分别由函数/、Q上的/*和g表示的目标图像、目标域、可能还有源域内所包含的图像数据。 Target image data collector 511 to collect the input image by the function in the herein are /, / g, and the Q * represents the target domain, there may be image data contained in the source domain. 混和计算模块512接收目标域定义、目标图像数据、可能还有源区域信息。 Blending computation module 512 receives the destination domain definition, the destination image data, and possibly the source region information. 此外,混和计 In addition, mixed meter

算模块512使用这些输入,基于源自或取自存储在内插指导数据存储515中的信息的内插指导,来计算混和的结果图像516的图像数据。 Calculation module 512 uses these inputs, the interpolation direction data taken from a memory based on or derived from information stored within the inner plug guide 515, image data to calculate the blended result image 516. 输出模块514接收混和的结果图像516的图像数据,该模块在显示器上呈现混和的结果图像516, 或者将该图像输出到存储设备或另一个应用程序。 Output module 514 receives the blended result image 516 of the image data, the module presents blended result image 516 on a display, or outputs the image to a storage device or another application.

图6描绘了本发明的一个实施例中混和的操作。 6 depicts an embodiment of the present invention, the mixing operation embodiment. 接收操作600接收对源(例如,源区域)的选择。 Receiving operation 600 receives the source (e.g., source region) is selected. 另一个接收操作602接收目标(例如,目标区域)的选择。 Another receiving operation 602 receives a target (e.g., a target region) selection. 定义操作604基于对源的选择来定义源域。 604 define the operation based on the selection of the source to define the source domain. 另一个定义操作606基于对目标的选择来定义目标域。 Another definition of the target based on the selection operation 606 to define the target domain.

收集操作608收集在本文中分别由函数/*、 Q上的/^和g表示的目标图像、目标域、可能还有源域内所包含的图像数据。 Collection operation target image 608 are collected herein / *, / and g ^ Q is represented by the function of the target domain, there may be image data contained in the source domain. 提供操作610通过诸如访问数据存储或提供硬编码的算法,提供受指导的内插器。 Operation 610 provides access to data by an algorithm, such as a store or provide hard-coded, to provide guided by the interpolator. 计算操作612从所收集的图像数据和受指导的内插器计算混和的结果图像。 Calculating operation calculates blended result image 612 from the image data collected and guided by the interpolator. 输出操作614将混和的结果图像输出到显示器或其它输出设备。 Output operation blended result image 614 is output to a display or other output device.

图7描绘了本发明的一个实施例中源自透明内插指导的混和结果。 7 depicts one embodiment of the present invention is derived from the result of mixing the transparent interpolation guide. 使用过滤内插指导尸将源图像700混和到目标图像702中,以产生混和的结果图像704。 Results interpolation guide corpse within the source image 700 using the filter to the target image blending 702, 704 to produce a blended image. 然而,从结果图像704中显而易见的是,单词"BLEND"的各个字母之间以及内部的间隔未被实际地呈现。 However, it is apparent from the results of the image is 704, and the spacing inside the actual presents is not between the word "BLEND" of individual letters. 相反,应用了称作透明的第二内插指导,来产生混和的结果图像706。 In contrast, the application of a second interpolation guide called transparent, to produce blended result image 706. 透明的算法在上表中给出。 Transparent algorithm is given in the table.

透明算法的替换表示是: Replace transparent representation of the algorithm is:

透明内插指导T在目标区域中的每个位置选择最大的图像梯度(从源图像数据或者从目标图像数据中选择)。 A transparent interpolation guide T each position in the target area of ​​the maximum image gradient of the selected (or select from the source image data from the image data of the target). 以此方式,混和的结果图像包含了与目标域中给定位置的源区域或者目标区域最明显的特征最匹配的图像数据。 In this manner, the blended result image comprises image data and the most significant feature of the target domain given location or target area of ​​the source region best match.

图8描绘了本发明的一个实施例中源自掩模内插指导#的混和结果。 FIG 8 depicts an embodiment the blended result derived from the interpolation # guiding a mask embodiment of the present invention. 在掩模内插指导中,将梯度最优化以满足预定标准,包括但不限于,低于阈值、高于阈值、在给定范围内、以及满足给定目标。 Inserted in the guiding mask, the gradient will be optimized to meet the predetermined criteria, including but not limited to, below a threshold, above a threshold, within a given range, and satisfies a given target.

如上所示,掩模内插指导#过滤目标图像数据,从而在混和的结果图像的目标区域中仅生成最显著的特征。 As indicated above, the masking interpolation guide # filtered target image data to generate only the most significant features in the target region of the blended result image. 例如,使用如前表中所示的掩模内插指导私对低(或者,替换地,高)梯度区域减弱源图像800的源区域804中的图像数 For example, using a mask as shown in the preceding table private low insertion guide (or, alternatively, high) gradient regions reduced number of source image the image source 800 in region 804

▽g(",其它据亮度,以产生混和的结果图像802中所示的类似卡通的效果。 替换地,可修改掩模内插指导i/来提供带通过滤器范围: . ▽ g ( ", according to other similar result image 802 shown in cartoon brightness to produce the effect of mixing Alternatively, the mask may be modified interpolation guidance i / range band-pass filter is provided:

<formula>formula see original document page 13</formula> <Formula> formula see original document page 13 </ formula>

图9示出了本发明的一个实施例中混和到源域维度以外的目标图像中。 FIG 9 shows an embodiment of the present invention is blended into the destination image outside the dimensions of the source domain embodiment. 示出了具有被目标边界3Q 902界定的目标区域的目标图像900。 900 shows a target image having target area 902 defined objectives of the boundary 3Q. 混和区域T 906 在内部由混和边界5T 904界定,在外部由目标边界3Q 902界定。 Mixing region T 906 is bounded by boundary mixture 5T 904 internally, externally delimited by object boundary 3Q 902. 可提供受指导的内插器,以在目标边界3Q 902和混和边界OT 904之间的区域T中提供一定量 Guidance may be provided by the interpolator, to the boundary region between the target T and blending boundary 3Q 902 OT 904 provided in an amount

的混和: A mixture of:

mill,贈-/*|2, 它具有边界条件/^^y^L。 mill, gift - / * | 2, which has a boundary condition / ^^ y ^ L. 此区域中混和效果的大小可由加权参数a控制。 Confounding effect of the size of this region may be a weighting parameter control.

用于实现本发明的图10的示例性硬件和操作环境包括计算机20形式的通用计算设备,包括处理单元21、系统存储器22、和用于将包括系统存储器在内的各种系统组件耦合到处理单元21的系统总线23。 For implementing an exemplary hardware and operating environment of FIG 10 according to the present invention includes 20 general purpose computing device in the form of a computer, including a processing unit 21, a system memory 22, and for the various system components including the system memory is coupled to the processing unit 21. the system bus 23. 可以仅有一个或有一个以上处理单元21,从而计算机20的处理器包含单个中央处理单元(CPU),或通常称作并行处理环境的多个处理单元。 There may only be one or more processing units 21, such that the processor of computer 20 comprises a single central processing unit (CPU), or more commonly referred to as a parallel processing environment of the processing units. 计算机20可以是常规计算机、分布式计算机、或者任何其它类型的计算机;本发明不受此类限制。 The computer 20 may be a conventional computer, a distributed computer, or any other type of computer; the present invention is not such a limitation.

系统总线23可以是各种总线结构中的任意一种,包括存储器总线或存储器控制器、外围总线、以及使用各种总线体系结构的任一种的局部总线。 The system bus 23 may be any of a variety of bus structures, including a memory bus or memory controller, a peripheral bus, and a bus using any of a variety of local bus architectures. 系统存储器也可简称为存储器,并且包括只读存储器(ROM) 24和随机存取存储器(RAM) 25。 The system memory may also be referred to as a memory, and includes read only memory (ROM) 24 and random access memory (RAM) 25. 包含诸如在启动时帮助在个人计算机20内部各元件间传递信息的基本例程的基本输入/输出系统(BIOS) 26储存在ROM 24中。 Such as a basic input contains the basic routines that help to transfer information between elements 20 inside a personal computer / output system (BIOS) 26 stored in the ROM 24 at startup. 计算机20还包括用来读或写硬盘(未示出)的硬盘驱动器27、用来读或写可移动磁盘29的磁盘驱动器28、 和用来读或写诸如CD ROM或其它光介质等可移动光盘31的光盘驱动器30。 The computer 20 further includes a read or write to a hard disk (not shown) of the hard disk drive 27 for reading from or writing to a removable disk drive 29, magnetic disk 28, and for reading or writing, such as a CD ROM or other optical media removable an optical disk drive 3031 of the optical disc.

硬盘驱动器27、磁盘驱动器28和光盘驱动器30分别由硬盘驱动器接口32、磁盘驱动器接口33和光盘驱动器接口34连到系统总线23。 Hard disk drive 27, magnetic disk drive 28, and optical disk drive 30 are a hard disk drive interface 32, magnetic disk drive interface 33, and an optical disk drive interface 34 is connected to the system bus 23. 这些驱动器及其相关联的计算机可读介质为计算机20提供计算机可读指令、数据结构、程序模块及其它数据的非易失性存储。 The drives and their associated computer-readable medium having computer-readable instructions providing a computer 20, a nonvolatile memory data structures, program modules and other data for the. 本领域技术人员应当理解,诸如磁带盒、 闪存卡、数字视频盘、贝努利盒式磁带、随机存取存储器(RAM)、只读存储器(ROM)等任何可储存能由计算机访问的数据的其它类型的计算机可读介质也 It should be understood by those skilled in the art, such as magnetic cassettes, flash memory cards, digital video disks, Bernoulli cartridges, random access memory (RAM), a read only data memory (ROM) and any other store can be accessed by a computer other types of computer-readable media may also

可用于该示例性操作环境中。 It may be used in the exemplary operating environment.

包括操作系统35、 一个或多个应用程序36、其它程序模块37及程序数据38 在内的若干程序模块可储存在硬盘、磁盘29、光盘3K ROM 24或RAM 25上。 A plurality of program modules include an operating system 35, one or more application programs 36, other program modules 37 and program data stored in the hard disk 38 including, magnetic disk 29, optical disc or 3K ROM 24 RAM 25. 用户可通过诸如键盘40和定位设备42等输入设备来输入命令和信息到个人计算机20中。 User may enter commands and information into the personal computer 20 through input devices such as a keyboard 40 and pointing device 42 and the like. 其它输入设备(未示出)可包括话筒、操纵杆、游戏垫、圆盘式卫星天线、 扫描仪等等。 Other input devices (not shown) may include a microphone, joystick, game pad, satellite dish, scanner or the like. 这些及其它输入设备常通过耦合到系统总线的串行端口接口46连到处理单元21,但也可用诸如并行端口、游戏端口或通用串行总线(USB)等其它接口连接。 These and other input devices are often coupled to the system bus via a serial port interface 46 coupled to the processing unit 21, but may also be used such as a parallel port, game port or other interface a universal serial bus (USB) and the like. 监视器47或其它类型的显示设备也经由诸如视频适配器48等接口连到系统总线23。 A monitor 47 or other type of display device is also connected via other interfaces such as a video adapter 48 is connected to the system bus 23. 除监视器之外,计算机通常包括诸如扬声器及打印机等其它外围输出设备(未示出)。 In addition to the monitor, computers typically include other peripheral output devices (not shown) such as speakers and printers.

个人计算机20可用到诸如远程计算机49等一个或多个远程计算机的逻辑连接在联网环境中运行。 The personal computer 20 may be a remote computer to a remote computer logic 49, such as one or more other connection operate in a networked environment. 这些逻辑连接是由耦合到计算机20或是作为其一部分的通信设备实现的。 These logical connections are coupled to a computer 20 or as part of implementation of the communication device. 远程计算机49可以是另一计算机、服务器、路由器、网络PC、客户机、对等设备或其它普通网络节点,且通常包括上述相对于个人计算机20所描述的许多或全部元件,尽管在图10中只示出了记忆存储设备50。 The remote computer 49 may be another computer, a server, a router, a network PC, a client, a peer device or other common network node, and typically includes many or all of the above-described elements within the personal computer 20 as described, although in FIG. 10 It shows only a memory storage device 50. 图10中描绘的逻辑连接包括局域网(LAN) 51及广域网(WAN) 52。 Depicted in Figure 10. The logical connections include a local area network (LAN) 51 and a wide area network (WAN) 52. 此类网络环境常见于办公室、企业范围计算机网络、内联网及因特网,它们均为网络的类型。 Such networking environments are commonplace in offices, enterprise-wide computer networks, intranets and the Internet, they are the type of network.

当用于LAN网络环境中时,计算机20通过属于一类通信设备的网络接口或适配器53连到局域网51。 When used in a LAN networking environment, the computer 20 through the communication device belongs to a class of a network interface or adapter 53 is connected to the LAN 51. 当用于WAN网络环境中时,计算机20通常包括属于一类通信设备的调制解调器54或用于通过广域网52建立通信的任何其它装置。 When used in a WAN networking environment, the computer 20 typically includes a modem 54 belonging to a class of communications device, or any other means for establishing communications over the wide area network 52. 调制解调器54可以是内置或外置的,它经由串行端口接口46连到系统总线23。 The modem 54, which may be internal or external, is connected via a serial port interface 46 to the system bus 23. 在联网环境中,相对于个人计算机20所描绘的程序模块或其部分可储存在远程记忆存储设备中。 In a networked environment, with respect to the personal computer 20 depicted in program modules or portions thereof may be stored in the remote memory storage device. 应当理解,示出的网络连接是示例性的,也可使用在计算机间建立通信链路的其它装置和通信设备。 It should be appreciated that the network connections shown are exemplary and other means may be used, and a communication device establishing a communications link between the computers.

在本发明的一个实施例中,选择模块、域定义模块、混和计算模块、和输出模块可被合并为操作系统35、应用程序36、或者其它程序模块37的部分。 In one embodiment of the present invention, the selection module, domain definition module, a blending computation module, and output module may be incorporated as an operating system 35, application programs 36, other program modules 37, or part of. 内插指导数据存储和图像数据可存储为程序数据38。 The interpolation guide datastore and image data may be stored as program data 38.

本文中所描述的本发明的各个实施例被实现为一个或多个计算机系统中的逻辑步骤。 Various embodiments of the invention described herein are implemented as logical steps in one or more computer systems. 本发明的逻辑操作被实现为:(l)在一个或多个计算机系统中执行的处理器实现的步骤序列,以及(2)—个或多个计算机系统内部互连的机器模块。 Logical operations of the present invention are implemented as: (l) a processor executing one or more computer systems to achieve the sequence of steps, and (2) - inside or more computer systems interconnected machine modules. 该实现是取决于实现本发明的计算机系统的性能要求的选择问题。 The implementation is a choice dependent on the performance of a computer system implementing the invention claims. 因此, 组成本文中所描述的本发明各实施例的逻辑操作在不同场合下被称作操作、步骤、对象或模块。 Accordingly, the logical operations of various embodiments of the present invention the composition described herein is referred to as operations, steps, objects, or modules at different occasions.

以上的说明书、示例和数据提供了本发明示例性实施例的结构和使用的完整描述。 The above specification, examples and data provide a complete description of the structure and use of exemplary embodiments of the present invention. 因为可以作出许多本发明的实施例而不会偏离本发明的精神和范畴, 因此本发明驻留于所附的权利要求书中。 Since many embodiments may be made of the invention without departing from the spirit and scope of the invention, the invention resides in the appended claims.

Claims (18)

1.一种从目标图像计算混和的结果图像的方法,所述方法包括: 在所述目标图像中定义具有边界的目标域, 提供一受指导的内插器,它包括内插指导和与所述目标域的边界相关联的边界条件;以及基于所述受指导的内插器计算所述混和的结果图像,以满足所述目标域的边界处的边界条件,并将所述内插指导与所述混和的结果图像的梯度之间在所述目标域内的差值最小化。 A blended result image calculated from the target image, the method comprising: a target domain having a boundary defining the target in the image, by providing a guided interpolator including an interpolation guide and with the boundary condition boundary of said associated target domain; and computing the blended result image based on said subject guided interpolator to satisfy the boundary condition at the boundary of the target domain, and the interpolation guide and between a gradient of the blended result image within the target difference is minimized.
2. 如权利要求1所述的方法,其特征在于,所述提供操作包含: 提供一掩模内插指导作为内插指导,所述掩模内插指导包括所述目标域内的目标图像的梯度,其中所述目标图像的梯度满足一预定标准。 2. The method according to claim 1, wherein the providing operation comprises: providing a masking interpolation guide as the interpolation guide, the masking interpolation guide including a gradient of the target image in the target domain , wherein the gradient of the target image satisfies a predetermined criterion.
3. 如权利要求1所述的方法,其特征在于,源图像的至少一部分被混和到所述目标图像中,以生成所述混和的结果图像,所述内插指导是所述源图像的函数。 3. The method according to claim 1, wherein at least a portion of the source image is blended into the destination image to generate the blended result image, the interpolation guide is a function of the source image .
4. 如权利要求1所述的方法,其特征在于,源图像的至少一部分被混和到所述目标图像中,以生成所述混和的结果图像,并且所述方法还包含:在提供-受指导的内插器之前,在所述源图像中定义具有边界的源域。 4. The method according to claim 1, wherein at least a portion of the source image is blended into the destination image to generate the blended result image, and the method further comprises: providing - guided by before the interposer is defined in the source domain having a boundary in the source image.
5. 如权利要求4所述的方法,其特征在于,所述提供操作包含: 提供一过滤内插指导作为内插指导,所述过滤内插指导包括所述源域内的源图像的梯度。 5. The method according to claim 4, wherein the providing operation comprises: providing a guidance interpolating the interpolation filter as a guide within the filter interpolation guide includes a gradient of the source image in the source domain.
6. 如权利要求4所述的方法,其特征在于,所述提供操作包含: 提供一透明内插指导作为内插指导,所述透明内插指导包括所述源域内的源图像的梯度和所述目标域内的目标图像的梯度之间的较大者。 6. The method according to claim 4, wherein the providing operation comprises: providing a transparent interpolation guide as the interpolation guide, the transparent interpolation guide including a gradient of the source image within the source domain and the the larger the gradient between the target image of said target domain.
7. 如权利要求1所述的方法,其特征在于,所述边界条件要求所述混和的结果图像中的图像数据在所述目标域的边界处的每个值等于所述目标图像中图像数据在所述边界处的对应值。 7. The method according to claim 1, wherein the boundary condition requires that the blended result image data at each image value at the boundary of the target region is the image data of the target image the corresponding value at the boundary.
8. 如权利要求1所述的方法,其特征在于,所述受指导的内插器由以下函数最小化表示:<formula>formula see original document page 3</formula>它具有边界条件/|5。 8. The method according to claim 1, characterized in that, guided by the interpolator is represented by minimizing the following function: <formula> formula see original document page 3 </ formula> having a boundary condition / | 5 . =/*|5。 = / * | 5. ,其中v表示内插指导,/表示所述目标域内的混和的结果图像,/*表示目标图像。 , Where v represents an interpolation guide, / represents the blended result image within the target, / * indicates the target image.
9. 一种用于从目标图像计算混和的结果图像的系统,所述系统包含: 域定义模块,它在所述目标图像中定义具有边界的目标域, 内插指导数据存储,它存储一受指导的内插器,所述内插器包括内插指导和与所述目标域的边界相关联的边界条件;以及混和计算模块,它基于所述受指导的内插器计算所述混和的结果图像,以满足所述目标域的边界处的边界条件,并将所述内插指导与所述混和的结果图像的梯度之间在所述目标域内的差值最小化。 9. A method for calculating blended result image from a target image, the system comprising: a domain definition module, which defines the target domain having a boundary in the target image, the interpolation guide datastore that stores a subject guided interpolator, the interpolator including an interpolation guide and a boundary condition of the target domain associated boundary; and blending computation module, the result of mixing the receiving its guidance based on said interpolator boundary condition at the boundary of the image, in order to meet the target domain, and the interpolation guide and the minimum of the difference between the gradient of the target domain blended result image.
10. 如权利要求9所述的系统,其特征在于,所述内插指导包括所述目标域内的目标图像的梯度。 10. The system according to claim 9, wherein the interpolation guide includes a target image of the target region of the gradient.
11. 如权利要求9所述的系统,其特征在于,所述内插指导等于0。 11. The system according to claim 9, characterized in that the inner guide is inserted is equal to 0.
12. 如权利要求9所述的系统,其特征在于,所述内插指导包括所述目标域内的目标图像的梯度,其中,所述目标图像的梯度满足一预定标准。 12. The system according to claim 9, wherein the interpolation guide includes a gradient of the target domain of the target image, wherein a gradient of said target image satisfies a predetermined criterion.
13. 如权利要求9所述的系统,其特征在于,源图像的至少一部分被混和到所述目标图像中,以生成所述混和的结果图像,所述内插指导是所述源图像的函数。 13. The system according to claim 9, wherein at least a portion of the source image is blended into the destination image to generate the blended result image, the interpolation guide is a function of the source image .
14. 如权利要求9所述的系统,其特征在于,源图像的至少一部分被混和到所述目标图像中,以生成所述混和的结果图像,并且所述系统还包含:在提供一受指导的内插器之前,在所述源图像中定义具有边界的源域。 14. The system according to claim 9, wherein at least a portion of the source image is blended into the destination image to generate the blended result image, and the system further comprising: providing a guide receiving before the interposer is defined in the source domain having a boundary in the source image.
15. 如权利要求14所述的系统,其特征在于,所述内插指导包括所述源域内的源图像的梯度。 15. The system according to claim 14, wherein the interpolation guide includes a gradient of the source image in the source domain.
16. 如权利要求14所述的系统,其特征在于,所述内插指导包括所述源域内的源图像的梯度和所述目标域内的目标图像的梯度之间的较大者。 16. The system according to claim 14, wherein the interpolation guide includes a greater of a gradient of the gradient of the target image and the target image in the source domain to the source domain.
17. 如权利要求9所述的系统,其特征在于,所述边界条件要求所述混和的结果图像中的图像数据在所述目标域的边界处的每个值等于所述目标图像中的图像数据在所述边界处的对应值。 17. The system according to claim 9, wherein the boundary condition requires that the blended result image data at each image value at the boundary of the target region is the target image in the image data corresponding to the value at the boundary.
18. 如权利要求9所述的系统,其特征在于,所述受指导的内插器由以下函数最小化表示:rnin, j"|JY/"-v| ,它具有边界条件几。 18. The system according to claim 9, characterized in that, guided by the interpolator is represented by the following function is minimized: rnin, j "| JY /" - v |, which has several boundary conditions. =/*",其中v表示内插指导,/表示所述目标域内的混和的结果图像,/*表示目标图像。 = / * ", Where v represents an interpolation guide, / represents the blended result image within the target, / * indicates the target image.
CN 200480003010 2003-02-25 2004-01-07 Image blending by guided interpolation CN100386772C (en)

Priority Applications (3)

Application Number Priority Date Filing Date Title
US45007803P true 2003-02-25 2003-02-25
US60/450,078 2003-02-25
US10/417,937 2003-04-16

Publications (2)

Publication Number Publication Date
CN1745386A CN1745386A (en) 2006-03-08
CN100386772C true CN100386772C (en) 2008-05-07

Family

ID=36140016

Family Applications (1)

Application Number Title Priority Date Filing Date
CN 200480003010 CN100386772C (en) 2003-02-25 2004-01-07 Image blending by guided interpolation

Country Status (1)

Country Link
CN (1) CN100386772C (en)

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101631189B (en) 2008-07-15 2012-08-29 鸿富锦精密工业(深圳)有限公司 Image synthesis system and method
CN101655969B (en) 2008-08-19 2011-12-21 方正国际软件(北京)有限公司 A method for editing a color gradient property
EP2284800B1 (en) 2009-07-23 2018-09-05 Samsung Electronics Co., Ltd. Method and system for creating an image
CN106339997A (en) * 2015-07-09 2017-01-18 株式会社理光 Image fusion method, device and system

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4602286A (en) 1982-01-15 1986-07-22 Quantel Limited Video processing for composite images
US5022085A (en) 1990-05-29 1991-06-04 Eastman Kodak Company Neighborhood-based merging of image data
US5870103A (en) 1996-09-25 1999-02-09 Eastman Kodak Company Method for creating realistic-looking composite images
US6081615A (en) 1997-06-24 2000-06-27 Kabushiki Kaisha Sega Enterprises Image-processing device and method of image-processing
US6094511A (en) 1996-07-31 2000-07-25 Canon Kabushiki Kaisha Image filtering method and apparatus with interpolation according to mapping function to produce final image

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4602286A (en) 1982-01-15 1986-07-22 Quantel Limited Video processing for composite images
US5022085A (en) 1990-05-29 1991-06-04 Eastman Kodak Company Neighborhood-based merging of image data
US6094511A (en) 1996-07-31 2000-07-25 Canon Kabushiki Kaisha Image filtering method and apparatus with interpolation according to mapping function to produce final image
US5870103A (en) 1996-09-25 1999-02-09 Eastman Kodak Company Method for creating realistic-looking composite images
US6081615A (en) 1997-06-24 2000-06-27 Kabushiki Kaisha Sega Enterprises Image-processing device and method of image-processing

Also Published As

Publication number Publication date
CN1745386A (en) 2006-03-08

Similar Documents

Publication Publication Date Title
Zhang et al. Parallax-tolerant image stitching
Li et al. Lazy snapping
US8520090B2 (en) Methods and apparatuses for image processing
Merritt et al. [26] Raster3D: Photorealistic molecular graphics
US9135732B2 (en) Object-level image editing
US7538908B2 (en) Image data processing apparatus, medium recording image data set, medium recording image data processing program and image data processing method
EP1372109B1 (en) Method and system for enhancing portrait images
Pritch et al. Shift-map image editing
US20050008254A1 (en) Image generation from plurality of images
JP4610043B2 (en) Mixing of colors when the transparency Consequences are present
US9058699B2 (en) System and method for image composition using non-destructive editing model and fast gradient solver
US7120293B2 (en) Interactive images
US20050134946A1 (en) Apparatus, method, and program for editing images
US6631240B1 (en) Multiresolution video
US7580566B2 (en) Image processing device, image processing method, storage medium, and program
US6757425B2 (en) Processing image data to transform color volumes
US20170270648A1 (en) Method for dynamic range editing
US20060104542A1 (en) Image tapestry
CN101589613B (en) A method and apparatus based on the adjustment layer and panoramas editor
JP2965119B2 (en) An image processing system and image processing method
US8823726B2 (en) Color balance
Agarwala et al. Interactive digital photomontage
US7206445B2 (en) Method, apparatus and recording medium for image processing
US7239328B2 (en) Lossless manipulation of media objects
US7519907B2 (en) System and method for image editing using an image stack

Legal Events

Date Code Title Description
C06 Publication
C10 Entry into substantive examination
C14 Grant of patent or utility model
C41 Transfer of patent application or patent right or utility model
ASS Succession or assignment of patent right

Owner name: MICROSOFT TECHNOLOGY LICENSING LLC

Free format text: FORMER OWNER: MICROSOFT CORP.

Effective date: 20150521

CF01