TW201137788A - Computerized method and system for pulling keys from a plurality of color segmented images, computerized method and system for auto-stereoscopic interpolation, and computer program product - Google Patents

Abstract

Described are computer-based methods and apparatuses, including computer program products, for three dimensional image rendering. In some embodiments, data indicative of a two dimensional image is stored in a data storage device, the two dimensional image comprising a plurality of pixels. A plurality of color segmented frames are generated based on the two dimensional image, wherein each color segmented frame comprises one or more objects. For each of the color segmented frames, a key is generated based on the one or more objects. A depth map is calculated for the two dimensional image based on the keys, wherein the depth map comprises data indicative of three dimensional information for each pixel of the two dimensional image. In some embodiments, a first two dimensional image and a second two dimensional image are received. A reduced pixel image is generated for each of the first and second two dimensional images, wherein each reduced pixel image comprises a reduced pixel size that is less than the original pixel size. Boundary information is calculated for each of the first and second two dimensional images. A depth map is calculated for the first and second reduced pixel images, wherein the depth map comprises data indicative of three dimensional information for one or more objects in the first and second reduced pixel images. A depth map is calculated for the first and second two dimensional images based on the boundary information for each of the first and second two dimensional images and the depth map of the first and second reduced pixel images.

Description

201137788 VI. Description of the Invention: [Technical Field] The present invention relates to a computer-based method and apparatus for three-dimensional reproduction, including computer program products, and in particular, for self-color segmentation image pull Take the base color gamut and autostereoscopic interpolation. [Prior Art] One-dimensional (3D) imaging is a technique of generating a depth illusion in an image to be perceived by a viewer. With stereoscopic imaging, a depth illusion (e.g., 'for two-dimensional (2D) images, photography, or movies) can be produced by presenting a slightly different image to each eye for a scene depicted within the medium. Typically, in order for the viewer to perceive the depth of the media, the user must view the stereoscopic images through some type of special viewing device, such as a special helmet or glasses. Autostereoscopic imaging is a technique for displaying a 3D image that can be viewed without using any particular viewing device, as compared to stereoscopic viewing. Although the media industry has made advances in imaging, there are still many challenges in effectively and accurately extracting objects from images and properly generating them. Color segmentation can be used to extract large homo-regional-deductive processes based on color and/or texture. Color segmentation acquires an original 2D image that can have hundreds or thousands of colors and reduces the number of colors in the image to a smaller subset of different colors. The resulting color segmented image can be used to generate a depth image representative of the depth information for each pixel or object within the image. In addition, in order to increase the calculation time required to generate 3D images from 2D images by 152792.doc 201137788, the solution is to use automatic scanning (r〇t〇sc〇ping). Transcoding system The process used to draw objects within an image. In the most traditional form of translating, the mapping is about producing a matte for one element on a real film (which is used to combine two or more image elements into a single most Z image) to Combining the element h over the other background may employ a Bezier eurves to present the 2D wheel of the object by evaluating an object at several points that are spaced apart differently and then converting the approximate order of the linear segments. The temple comes to the automatic line ^ line. ^, by only ❹ (four) curve 'because of limiting the number of Bayes points to increase the efficiency of 3 〇 reproduction efficiency (for example, the fewer points required for frame-by-frame transposition, the less processing time due to the fewer points that need to be manipulated Fast) and usually only "loosely" outlines the object. Therefore, it is advantageous to use as few Bates points as possible, which results in only a rough outline of the trace. A depth map can be generated to indicate which of the object regions within the image of the buckle are closer to the viewer or further away from the viewer. The tube can generate a depth map based on a single image, but typically uses a stereo pair to generate a depth map (by one For the corresponding camera-to-image, the cameras are configured so that each camera captures two different vantage points and there is a “know-closed” between the cameras. To generate a depth map, it is usually required to make the image The inter-common point is correctly associated with the vantage point information. For example, several segments can be compared between images (eg, a predetermined pixel square) to determine commonalities. However, this process is extremely dependent on the commonality. The accuracy of the selection within the two images 'this selection is an important and time consuming choice. Therefore, although various techniques have been adapted to produce 3D images to increase efficiency and speed, the process is still 152792.doc -6 - 201137788 Time, complexity and expensive. [Invention] In the aspect of the invention, the invention is characterized by a computerized method for extracting the primary color gamut from a plurality of color segmentation images. The method includes storing data indicating the two-dimensional image in a data storage device, the two-dimensional image comprising a plurality of pixels, the method further comprising generating a plurality of colors based on the one-dimensional image by a color segmentation unit of a computer a segmentation frame, wherein each color segmentation frame includes one or more objects. The method further includes: based on the one or more of each of the color segmentation frames by the color segmentation unit The object generates a primary color gamut; and the depth map unit of the computer calculates a depth map for the two two-dimensional images based on the primary color gamut, wherein the depth map includes three-dimensional information indicating each pixel in the two-dimensional image. In another aspect, the present invention comprises a system for self-complexing a color segmentation scene "image capture color gamut. The system includes a data storage device configured to store a two-dimensional image. Data, the two-dimensional image comprising a plurality of pixels. The system includes a color segmentation unit responsive to the data storage device configured to generate a plurality of colors based on the two-dimensional image A color segment frame 'which causes each color segment frame to include one or more objects, and a color gamut is generated based on the one or more objects for each of the color segment frames in the section. The system includes a depth map unit coupled to the color segmentation unit and the data storage device, configured to calculate a depth map based on the primary color gamut images, such that the depth is '' The data of the three-dimensional information of the mother pixel of the two-dimensional image. 152792.doc 201137788 In another aspect, the invention comprises a computer program product tangibly embodied in a computer readable storage medium. The computer program product is operative to cause the data processing device to execute the following instructions: storing data indicative of the two-dimensional image in the (4) storage device towel, the two-dimensional image comprising a plurality of pixels; and color segmentation by a computer The unit generates a plurality of color segmentation frames based on the two-dimensional image, wherein each color segmentation frame includes - or a plurality of objects. The computer program product also includes an operation operative to cause the data processing device to: generate, by the color segmentation unit, a primary color gamut based on the one or more objects for each of the color segmentation frames; & The depth map of the computer is calculated for the two-dimensional image based on the base color gamut - the depth map includes data indicating three-dimensional information of each pixel in the two-dimensional image. In another aspect, the invention includes a method of computerization for autostereoscopic interpolation. The method includes: receiving, by the computer-input unit, a first-two-dimensional image and a second two-dimensional image, each of the two-dimensional images including a pixel size; and the pre-processing unit of the computer Each of the two-dimensional image and the second two-dimensional image produces a reduced pixel image, wherein each-reduced pixel image includes a smaller than the pixel size - a reduced pixel size. The method also includes: calculating, by the pre-processing unit, boundary information for each of the second-dimensional image and the second two-dimensional image; wherein the computer-depth image unit is for the first-reduced pixel image And the third reduced pixel image calculation-depth map, wherein the depth map includes data indicating three-dimensional information of the first-reduced pixel image and the second reduced pixel image towel or the plurality of objects; And the depth map unit of the computer is based on the first two-dimensional I52792.doc 201137788 image and the second two-dimensional image order, Λ /A ± boundary boundary information and the first reduced pixel image and the The depth map of the second reduced pixel image is calculated for the first two-dimensional image and the second two-dimensional image. In another aspect, the invention includes for autostereoscopic interpolation

System. The system includes an input unit configured to receive a first-two-dimensional image and a second I--image. Each of the two-dimensional images includes a pixel size. The continuation system also includes communicating with the input unit, u. ^ ^ pre-processing, configured to generate a decrease for each of the first two-dimensional image and the second two-dimensional image a pixel image in which a small pixel image includes less than one of the pixel sizes, such as a pixel size, and for each of the first two-dimensional image and the "first-dimensional image" Calculating boundary information, the system includes a depth map unit in communication with the pre-processing unit, wherein the depth map is configured to calculate a depth map for the first-reduced pixel image and the second reduced pixel image, wherein the depth The figure includes data indicating three-dimensional information of the first-reduced pixel image and the first-reduced image (four) image or the plurality of objects; and based on the first-two-dimensional image and the second two-dimensional image The boundary information of each of the boundary information and the depth map of the first reduced pixel image and the second reduced pixel image are used to calculate a depth map for the (four)-two-dimensional image and the (four) two-dimensional image. The invention comprises a computer program product. The product is tangibly embodied in a computer readable storage medium. The computer program product includes an operation operative to indicate that the data processing device executes the following instructions: a receive-first-two-dimensional image and a second two-dimensional image each-two-dimensional The image includes a pixel size '· and a reduced pixel image is generated for each of the first two-dimensional image and the second two-dimensional image 152792.doc 201137788, wherein each reduced pixel image includes less than the Pixel size - reducing the pixel size. The computer program product also includes instructions operable to instruct a data processing device to perform the following operations: calculating boundary information for each of the first two-dimensional image and the second two-dimensional image Calculating a depth map for the first reduced pixel image and the second reduced pixel image, wherein the depth map includes one of the first reduced pixel image and the second reduced pixel image Or three-dimensional information of the plurality of objects; and boundary information based on each of the first two-dimensional image and the second two-dimensional image and the first reduced pixel image and the second subtraction The depth map of the pixel image is used to calculate a depth map for the first two-dimensional image and the second two-dimensional image. ^ Other real money, any of the above-mentioned aspects can include one or more of the following features. In some instances, the systems and methods may allow adjustment of the Bezier point of an object--Bez plot (eg, after the initial color segmentation is not accurate enough or does not produce the desired effect). In an example, for each color segment frame, one edge of each object of the one or more object orders may be defined by automatically calculating one of the object phases, the Bates diagram including a plurality of Bezier points of the object and a plurality of Bezier curves connecting the plurality of Bezier points; receiving the indication to adjust the information of the Bezier diagram' and generating a detailed picture of the object based on the data, wherein the detailed map of the money includes A plurality of additional points combined with the plurality of (four) points. In some instances, generating the equal color & each knife frame may include defining a plurality of basic colors, wherein each element is used to generate a color segmentation frame.

Each of the plurality of basic colors - the A basic color contains a predefined color 152792.doc 201137788 value range, including pixels of one of the color values within the predefined color value range and based on the basic color A color segmentation frame is associated. Information indicative of a new color value of one or more of the plurality of basic colors may be received, and a predefined range of color values may be adjusted for each of the one or more colors based on the material. For each of the plurality of basic colors, a plurality of basic colors may include a color pair, the color pair and the dark color. The plurality of basic colors may include brown and beige. In the example of the method, the calculating the depth map may include: determining that the two-dimensional representation of the object in the two-dimensional image is not in the field of view in the two-dimensional image: one part of the object enters the field of view; and stretching the The side of the back/object of the object or one of the P-cuts of the object filled in the field of view may include: determining that one of the two-dimensional images does not result in The portion of the object in the field of view that disappears within the field of view in the field of view; and shrinks one side of the object to conceal the portion of the object that disappears within the field of view. In a κ example, calculating the depth map may include calculating three-dimensional information for each pixel based on one of the key factors of the plurality of color binning frames. The key factor:: HLS (hue, brightness and saturation) color space containing the pixel: ▲Differential can include the level of the determination of saturation, the level of brightness or both: to the level of the I low level; And assigning a near-depth value to the pixel in the case of the level of saturation or brightness: or in the case of the saturation=brightness level-low level, assigning a far depth value to The one of the key factors may include a value change of 152792.doc • 11 - 201137788 among the pixel and the adjacent pixel group. The calculation can be sentenced to eight. No system - part of the plane two " change to determine that the pixel is in the part of the plane of the pixel, privately send a near depth value to the pixel. In the case of other instances, # Μ μ m _ i section 3 hai key factor can be packaged in one or more of the objects in each of the knives in the box. Calculate the object - the position of the object, then, the brother IT 4t ·. The lower position of one of the frames of the Bachelor knife is the position of the object in the lower part of the position. In the case of an upper position of the ^ & frame, assigning a pass depth value to the parent pixel in the object, where the position of the object is in the corresponding object of another color segmented circle frame, Assigning a near-ice value to a pixel in the object, or at the location of the object

知知刀#又框框——End A 兮铷 兮铷 In the case of a knife, assign - far depth value to each child in the object to ^ pixels. The key factor may include a ratio of the size of one of the plurality of mothers in each color segment frame to the color segment', and the calculation includes for each object based on the two-two heart & *degree value for each pixel in the object. The key factor is the position of one of the objects in the image, and the position of the corresponding object in a continuous two (four) image. The calculating may include: 2 positioning the object in the two-dimensional image differently than the position of the corresponding object in the continuous two-dimensional image, and assigning a near depth value to each pixel in the object. In some examples, the data of a previously generated depth map can be stored in the data storage device. The edge ice level map includes two-dimensional information indicating a pixel corresponding to a two-dimensional image. Data, and may generate a new depth map based on the previously generated deep 152792.doc •12-201137788 degree map, wherein the new depth map includes a three-dimensional representation indicating a larger range of each pixel in a corresponding two-dimensional image Information on information. Calculating the depth map can include applying one or more experience rules to adjust the depth map, each rule configured to adjust the depth map based on human perception of one of the one or more objects in the two-dimensional image. In other examples, the system can include an input unit configured to: receive data indicative of the two-dimensional image and store the data in a data storage device; and receive information indicative of a previously generated depth map and The data indicating the previously generated depth map is stored in the data storage device. The system can include an edge generation unit configured to generate edge information for each of the two dimensional images. The system can include a three-dimensional experience unit configured to apply - or a plurality of experience rules to adjust the depth map, each-rule configured to be based on human perception of one or more of the objects in the two-dimensional image Adjust the depth map. / Α π w 丨 久 久 久 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 Including the first two-dimensional image of 5 Hai - like the second - the heart of the material * (four) points and 0 two-dimensional pixels - minus::: two::: the first - the reduction of the image, etc. Pixel image, the objects of the object _ degrees = based on the difference to generate the data. The pixel position of the test is in other examples, the method includes the two corresponding pixels * the Xth corresponding pixel, wherein the image is 'and the two corresponding pixels ^ are located at the 4th - reduced pixel shadow 152792. Doc—the second pixel is located in the pixel image of the second decrease-13.201137788; the calculation refers to the coherence, the side...the pixel is separated from the second pixel by more == and the assignment indicates the two corresponding pixels In the depth map, the material of one of the objects is away from the three-dimensional symplectic inclusion of the object. The method can further include: comparing two corresponding images two = two corresponding pixels - the first pixel is located First reduction = prime: in the image 'and the two corresponding pixels - the second pixel is located in; 1 small pixel image; the calculation indicates the first pixel and the second pixel ΐ = two values; and the assignment indication The two corresponding pixels are in the deep two-in-one to indicate that the data of the object is close to one of the two-dimensional views of the object. ^In an instance, the pixel size can be calculated as each corresponding two-dimensional In the shadow - the pixel length and the width of a pixel Calculating the depth map for the first two-dimensional image and the second two-dimensional image may include: calculating a depth map of each boundary pixel for the boundary information based on the first-dimensional image and the second two-dimensional image; Determining the depth information of the depth map of the remaining pixels by using the data of the corresponding boundary pixels of the pixel image and the second reduced pixel image. For other real money, calculating the boundary information includes generating a complex number based on the two-dimensional image. a color segmentation frame, wherein each color segment frame includes one or more objects; and based on the pixels of the first-reduced pixel image and the second reduced pixel image The color segmentation frame sets the pixel, - the boundary point indicator 'where the boundary point indicator includes information indicating whether the pixel is a boundary point. The verification can be performed for each of the plurality of color segmentation frames. Make sure it contains one or more internal polymer parts, each internal polymer part 152792.doc 14. 201137788

Including an identifiable boundary line D. In some examples, calculating the depth map includes identifying a hidden pixel of the first reduced pixel image by identifying a visible pixel within the first reduced pixel image, wherein The visible pixel has a corresponding pixel in the first: reduced pixel image. Generating a third two-dimensional image based on the depth map of the first two-dimensional image and the second two-dimensional image, wherein the third two-dimensional image includes a first advantageous point between the first two-dimensional image and An advantageous point between the second advantageous point of the second two-dimensional image, and wherein the third two-dimensional image comprises a region that enters one of the fields of view or disappears within the field of view based on the identified hidden pixel. In other examples, the system can include a color shirt segmentation unit in communication with the pre-processing unit configured to generate a plurality of color/segment frames based on the two-dimensional image, wherein each color segmentation frame includes One or more items. The system can include a conversion unit configured to generate a second two-dimensional image based on the depth map, wherein the third two-dimensional image includes a first vantage point between the first two-dimensional image and the first An advantageous point between the second advantageous point of one of the two two-dimensional images. The techniques described herein, including both methods and apparatus, may provide one or more of the following advantages. These techniques provide faster, more efficient, and more accurate three-dimensional conversions than existing conversion tools by generating an accurate depth map from a two-dimensional image and then converting the depth map into a three-dimensional image (e.g., via a stereo pair). By segmenting the two-dimensional image colors and identifying object boundaries based on the color segment frames, the system and method of the present invention achieves a significant time savings over the generation of boundary information by other means. In addition, initially, 152792.doc -15- 201137788 may be generated and adjusted by the user to generate saturation of each color of the color segmented images to customize the techniques according to a particular image (or frame) set. . Further, although the system provides for automatically generating a Bezier map for objects within the color segmentation frame, the Bezier point can also be manually adjusted (eg, the accuracy of the initial color segmentation is insufficient or cannot be produced) In the case of a desired three-dimensional effect). The system and method disclosed herein can operate not only on a two-dimensional image, but also on a previously generated depth map (eg, by using a larger custom depth range, the depth information contains more information to determine and / or distinguish between nearer and farther objects, which results in a more accurate depth map than a wide range of data points (c〇arsei plant grained) depth map). Further, human experience data can be used to more accurately convert 2D images (e.g., human experience data about facial features). In addition, the first one-dimensional image and the last two-dimensional image (or frame) of a frame sequence can be converted, and then the first two-dimensional image can be automatically converted based on the first one-dimensional image and the last two-dimensional image All 2D images between the image and the last 2D image. The last two-dimensional image of the sequence can then be set to a first two-dimensional image of a second two-dimensional image sequence, and the process repeated recursively. Additionally, if desired, the systems and methods disclosed herein can edit one of the two-dimensional image sequences to automatically convert the two-dimensional image. Other aspects and advantages of the present invention will be apparent from the description of the appended claims. The above and other objects, features and advantages of the present invention, as well as the present invention, will be more fully understood from the following description of the embodiments. In summary, the techniques disclosed herein provide automatic color segmentation for two-dimensional images (eg, several frames of an animation), identification of object boundaries based on such color segmentation frames (eg, a Bezier map) , which can be automatically adjusted and then fine-tuned), generate an accurate depth map from the 2D image (eg, have a custom depth range), and then convert the depth map into a 3D image (eg, via a stereo These techniques can operate on any type of two-dimensional image and can also operate on a previously generated depth map (eg, lacking one of the fine-tuned details used by the current technology). Figure 1 illustrates An exemplary 3D reproduction system 1 of the present invention includes a computer 1〇2. The computer 1〇2 includes a depth generator 104, an input unit 1〇6, and a conversion unit 1〇8. The input unit 1〇6 is in communication with the depth generator 104. The depth generator 1〇4 includes a pre-processing unit 110 and a database 112 (e.g., a data storage device). The depth generator 1〇4 also includes a color segmentation unit. 114 and a depth map creation unit 116. The color segmentation unit 114 includes an edge generation unit 118. The depth map creation unit 116 includes a 3D experience unit 120. Various components of the depth generator 1〇4 (eg, the pre-processing unit 110 to depth) The experience unit 12 of the map making unit 116 communicates with each other. The computer 102 is specifically configured to generate a 3D image computing system (eg, a programmable, processor-based system). Including, for example, a microprocessor, a hard disk drive (eg, database 12), random access memory (RAM), read only memory (R〇M), input/output (1/〇) 152792.doc •17· 201137788 Road and any other necessary Lei Cong έ 4 shame components. The computer 102 is preferably adapted for use with various types of storage devices (continuous and replaceable), such as (for example, portable disk drives, magnetic storage (such as floppy disk), solid state storage (such as - flash memory card), optical storage (such as - CD or CD) and / or network / Internet The network storage system can include - or more The brain 'includes (for example, a Windows/.. UNIX or other suitable operating system, an IBM/PC compatible computer) or one of the personal computers (for example, a workstation (such as a SUN or Silicon phics workstation) Preferably, the graphical user interface (gui) is included. The non-generator 104 can generate a depth map of the 2D image and/or process and refine the depth map generated by the prior art. For example, the depth generator 1〇 4 can be imported (eg 'via input unit 106, database 112 or via an internet or Ethernet connection (not shown)) - previous depth image and more accurately recalculate and fine tune the depth map The parameters. The depth generator (10) can also receive a 2D image that does not have an associated depth map and does not rely on any other additional data to calculate a depth map for the 2D image. The color segmentation unit 114 can generate a plurality of color segmentation images (eg, for a 2D image), and the depth map creation unit 116 can use the color segmentation images to generate the depth map. The input unit 106 enables communication of information to Depth generator 丨〇 4. For example, the wheeling unit 106 provides an interface for a user to communicate with the 3D rendering system 1 via the input device 122 (e.g., the input device 122 can transmit the data 124 to the input unit 106). The terms user and operator refer to one of the 3D reproduction systems and are used interchangeably. 152792.doc *18· 201137788 Input device 122 may include any device that enables the user to provide input to the computer. For example, the input device 122 can include a conventional input device such as a keyboard, a mouse, a trackball, a touch screen, a touchpad, a voice recognition hardware, a dial, a switch, a button, and a foot. A step switch, a remote control device, a scanner, a camera, a microphone, and/or a joystick. For example, the input unit 1G6 can be configured to receive a two-dimensional image from the input device 122. The conversion soap element 108 converts a depth map of a 2D image into an image. In some embodiments, the conversion unit 〇8 converts the _frame sequence (e.g., an animated sequence frame). For example, the depth generator 1〇4 can convert the first frame and the last frame of any given segment of a moving frame, and then automatically convert all of the between the first frame and the last frame. Frame. The last-frame can then be set to the first frame of a second sequence of frames. In addition, rightly, the depth generator 1〇4 can edit a frame between the first frame and the last frame of a sequence of frames. The edge generation unit 117 is configured to generate edge information (10) for each of the two-dimensional images, and the edge generation unit 118 rotates the objects within the image, and the color segmentation frame is rotated. Although the edge generation unit ι 8 is shown as a unit within the color segmentation unit m, this is for exemplary purposes only, and the edge generation unit may include one unit separate from the color segmentation unit ι 4 . The three-dimensional experience unit earns configuration to apply one or more experience rules for use in generating and/or adjusting a depth map (e.g., a depth map generated by depth map authoring unit 116). Each-rule is configured to be based on human perception 152792.doc •19· 201137788 Adjust the depth map. For example, the rules can be configured to adjust the depth map based on human perception of one of the one or more objects in the two-dimensional image. In some embodiments, when depth map authoring unit 116 calculates a depth map, depth map authoring unit 116 is configured to apply one or more experience rules to adjust the depth map. Advantageously, the 3D experience unit 120 can be configured to incorporate human experience material. Most 3D rendering programs do not take into account the properties of an image (eg, shadow, light angle, unintended color 4 4) or scenes (such as faces, celestial bodies, etc.) to produce a depth map. Apply the same rules to a specific image. Thus, in the case where a certain angle of light in a particular image projects a strange shadow, the programs typically interpret the object based on the colors and may render the object incorrectly due to the strange shadow. . This may not always produce a "correct" 3D image associated with a 3D image that a human viewer would perceive when viewing the same object. Compared to the rule of direct application based on pixel color, people do not always perceive an image or object based solely on color. Unlike the program, the human eye has an illusion that the person knows the depth even if the shadow or color is "incorrect" (for example, a nose always protrudes from the face, whether or not a shadow is projected on the nose to cause pixel information. Indicates that the nose is falling inward). Therefore, automatic 2d_3D conversion does not always result in a correct 3D image due to human illusions. In some instances the 'depth generator 104 can be configured with an r-face recognition illusion' that incorporates information about how a person perceives a face. For example, δ, as described above, the smear facial illusion verifies that the nose always protrudes from the face rather than inward in the face (for example, even if the shadow indicates the nose inward 152792.doc -20- 201137788 or other h Rules may be defined in the Bay Inspection Unit 12A to implement a facial recognition illusion. In some instances, additional colors may be added to the color segmentation unit 114 for use in segmenting the image. For example, due to Add color beige and brown to the number of shades in the face. Light-colored pairs and dark pairs can also be used for beige and brown to fine-tune the color of human anatomy. Advantageously, it can be initially used by the user. The saturation of each color is generated and adjusted to adjust the depth generator 104 according to a particular set of - or multiple images. This and other illusions can be built into the system to more accurately define the 2D-3D conversion process. The device 122 is in operative communication with the computer 102. For example, the input device 122 can be coupled to the computer 1〇2 via an interface (not shown). The interface can include a physical interface and/or a software interface. The interface can be any conventional interface such as, for example, a wired interface (eg, serial, USB, Ethernet, CAN bus and/or other cable communication interface) and/or a wireless interface (eg, Wireless Ethernet, wireless serial, infrared, and/or other wireless communication system) The software interface can reside on the computer 102 (eg, in the input unit 〇6). The display 126 is a computer 102 and uses A visual interface between the displays. The display 126 is coupled to the computer 102 and can be any suitable device for displaying text, images, graphics, 3D images, and/or other visual output. For example, the display 126 can include a standard. Display screen (eg, lcd, CRT, plasma, etc.), a touch screen, a portable display (9) such as glasses (g) or goggles (g0ggie) glasses protection (eyewea (1) a projection display, one-headed The display, a hologram display and/or any of its 152792.doc •21 - 201137788 his visual rims are located on the device. The display 126 can be placed on or near the computer 102 ('kp, gong+. Included in one of the cabinets of the computer 102) or remote from the computer 102 (for example, also u+.^.go's installed on one wall or other location suitable for viewing by the user). Turtle-. Any of the assets useful for a 3D rendering such as, for example, depth maps, color segmented images, stereoscopic images, autostereoscopic images, etc. are shown. Figure 2 illustrates a color segmentation derived from an image in accordance with the present invention. An exemplary image 200 is shown. The two-dimensional image 202 includes a background 204, a scene 2 (e.g., a table or cabinet), a first object 208, and a second object 210. The one-dimensional image 202 includes simplified objects for illustrative purposes, and the method and system can handle any image regardless of the complexity of the object within the scene/image. The drawing 200 includes four color segment frames 2丨2Α to 12D (collectively referred to as color segmentation frames 212). The color segmentation frame package 3 is for one of the backgrounds 204 color segmentation object 214. For example, the color used for the production knife and the shirt image 2 12A includes a range of color values (for example, RGB or HLS color space) of the entire moon hall, 204 capturing the two-dimensional image 2〇2, which is expressed as a Color segmentation object 2 14. Similarly, color segmentation frame 212B includes color segmentation object 216 for foreground 2〇6, color segmentation frame 21 2C includes color segmentation object 218 for first object 208, and color segmentation frame 212D includes A color segmentation object 220 for the second object 210. As will be explained below, depth generator 104 may automatically scan one-dimensional image 202 based on color segmentation frame 212. The primary color gamut can be pulled from each of the color segmentation frames 212. A primary color gamut represents the mother color used during the color segmentation process. For example, the green primary color gamut may represent a particular range of color values in the green chromatogram 152792.doc • 22- 201137788 (eg, for a tree with one green leaf, depending on the range of color values associated with the green primary color gamut, A specific portion of the greenish green leaf having the green value range is captured for the green primary color gamut). For example, the base color gamut of color segmentation frame 212D can represent the outline of color segmentation object 220, which is automatically rotated by depth generator ι4 to derive the primary color gamut. As will be explained in more detail below, the combination of the Betz points on the line around the color segmentation object 220 can be combined and then fine-tuned to make it more accurate to fit the color segmentation object 2 2 〇 ( If necessary, to translate the color segment object 220. In some instances, in the region being scanned, the user can adjust the depth information such that each object in the image can be individually calculated (eg, for the first object 2〇8 and the second object 21〇) The depth map of the image. Advantageously, the 2D image can be automatically rotated at the depth generator 丨〇4 when, for example, the accuracy of the initial color segment is insufficient or does not produce a desired effect (eg, an incorrect 3 〇 effect) Manually adjust the Bezier point of a Bezier after 2〇2. 3 illustrates an exemplary method 300 for pulling a base gamut from a color segmented image (e.g., 2D image 202 of FIG. 2, which is color segmented into color segmentation frame 212) in accordance with the present invention. At step 302, the depth generator 1〇4 stores data indicating the two-dimensional image (eg, the two-dimensional image 2〇2) in a data storage device (eg, the f library 112), the two-dimensional image including the plural Pixels. At step 〇4, depth generator 丨〇4 receives information to define a plurality of basic colors 'where each—a basic color is used to generate—a color segmentation frame for generating one of color segmentation frames 212) . At step 3〇6, the depth generator 104 generates a plurality of color segmentation frames based on the two-dimensional image (eg, by color segmentation I52792.doc-23·201137788 unit m), wherein each color segmentation frame includes One or more objects (eg, color segmentation frame 2丨2). At the step, the depth generator 104 selects a color segmentation frame 4 from the complex (four) color, step 31Q, and the depth generator ι4 generates for each of the color segmentation frames based on the one or more objects. (9) A primary color gamut such as 'by color segmentation unit. At step 312, the depth generator determines if there are any remaining color segmentation frames. If there is a remaining color segment frame, the method returns to the step. If there are no remaining color segmentation frames, then the method 300 continues to step 31 "At step Chuan" the depth generator 104 calculates three-dimensional information for each pixel based on one of the plurality of color segmentation frames. At step 316, the depth generator 104 calculates (eg, by the depth map generation unit 116) a depth map for the two-dimensional image based on the basic color gamuts, wherein the depth map includes a three-dimensional representation of each pixel of the five-dimensional one-dimensional image. Information about the information. In some embodiments, the depth generator (10) receives (10) the information indicating the two-dimensional image as the '(four) input unit 1G6) and stores the asset in the data storage device (eg, 'database m). In some implementations 2 the depth generator 104 can be configured to receive an indication of a previously generated

The information of St and the data indicating the previously generated depth map are stored in the far fat storage device. 'Step 304, the color segmentation unit can be configured to store a predetermined number of g Hr colors' wherein each of the predetermined colors is used to generate color bins and private. As described above with reference to the 3D experience unit 1〇2 of FIG. 1, the basic colors may include colors such as yellow, cyan, magenta, red, green, beige, brown 152792.doc • 24·201137788 color, blue, and the like. . Additionally, in some embodiments, color pairs may be used for each of the basic colors including one of a light color and a deep color for each of the basic colors (eg, 'such as 'dark red and light red, deep beige, and light beige and many more). For each basic color or color from a color pair (collectively referred to as color)', the color may include a predefined range of color values (e.g., RGB or HLS color space). These color ranges can be used when analyzing pixels to determine if a pixel color is within the color range. If a pixel of the image has a color value within the predefined color value range, the pixel is associated with a color segmentation frame generated based on the basic color, for example, assuming that the basic color "light red" has A range is r=128-255, G=0-95, B=〇-95, and the basic color "dark red" has a range of r=64_i27, G=0-95, B=0-95. Then, one of the color values having a range of R = 128_255, g = 0-95, and b = 〇_95 is associated with the basic color "light red". In the case where a region has a color value in the range of R=100-150, G=0-95, B=〇-95, the region is related to one of the basic colors of “light red” and “dark red”. Union. In some embodiments, the range of basic colors is automatically adjusted to cover the entire area corresponding to an item having a similar color. Depth generator 丨04 can be configured to allow adjustment of the range of color values associated with each elementary color (e.g., based on data received from a user). The color segmentation unit 114 can be configured to receive data indicative of a new color value range of one or more of the plurality of basic colors, and adjust the based on the data pin H or each of the plurality of colors A predefined range of color values. Advantageously, in the case where a particular image primarily contains a color (eg, various deep greens), the color values can be adjusted to distinguish between various shades of green 152792.doc •25·201137788 colors to cause color segmentation. Unit n4 produces a plurality of color segmentation frames. Otherwise, since the image is mainly composed of green shades, the color segmentation unit 114 can generate fewer color segment frames using preset values, and it is difficult to finely distinguish between various shades of green due to a large range of color values. Minute. With respect to step 306 and referring to FIG. 2, depth generator 1 〇4 generates a color segmentation frame 212 based on two-dimensional image 202. Each color segmentation frame 212 includes one or more objects (e.g., color segmentation image 212D includes color segmentation object 220). Regarding step 3 1 0 ' depth generator 104 generates a primary color gamut for each color segmentation frame based on one or more objects within the color segmentation frame. As mentioned above, one of the primary color gamuts of a color segmentation frame can represent an object within a particular color segment frame. Figure 4 illustrates an exemplary method for defining an edge of an object within an image in accordance with the present invention. At step 4〇2, the edge generation unit 11 selects a color segmentation frame from a plurality of color segmentation frames 丨2. At step 404, the edge generation unit selects an object (eg, color segmentation object 218 within color segmentation frame 212C) from one or more objects within the color segmentation frame ^ at step 4〇6 The edge generation unit 110 automatically calculates a Bezier map of the selected object. The Bezier diagram includes a plurality of Bezier points for the object and a plurality of Bezier curves connecting the plurality of Bezier points. The Bezier diagram can be used to generate an initial starting point for the edge of each object. At step 4〇8, the edge generation unit 118 receives the information indicating the adjustment of the Bezier map. For example, a user may desire to add additional points about an item to more accurately define the line around the item 152792.doc • 26· 201137788. At step 410, edge generation unit 118 generates a more detailed view of the object based on the material, wherein the more detailed map includes a plurality of additional points combined with the plurality of Bezier points. At step 412, edge generation unit 丨 18 determines if there are any remaining objects within the selected color segment frame. If edge generation unit ι 8 determines that there is an additional object, then method 400 continues with a return to step 404. If the edge generation unit 118 determines that there are no additional items in the selected color segment frame, the method 400 continues to step 414. At step 4M, the edge generation unit completes defining the edge of each object within the selected image. Method 4 is performed for each color segmentation frame from a plurality of color segmentation frames. Advantageously, in some embodiments, an initial bezial ffi ' is used for efficiency purposes and for a user to provide a starting point for the boundary of the object - and the user can subsequently fine tune the The edge map of the object. With regard to step 314, depth generator 1 4 calculates three-dimensional information for each pixel based on the key factor of the plurality of color segmentation frames. In general, the depth map is broken into black and white images, however the system and method of the present invention contains more data for each pixel by using a custom depth range. For example, 'a custom depth range 65 to 65, 28 〇 can be used for each pixel to fine-grain the depth information data point (fine_grain) (eg, one pixel to the farthest depth of the viewer) Value, but with a clothing = range 65, one of the 28 pixels is the closest depth value to the viewer). For example, conventional RGB values (e.g., from 〇 to 255) can be converted to a custom depth range by using a predetermined equation (e.g., configured by depth map authoring unit H6). Advantageously, by using a larger custom depth range, depth information 152792.doc -27· 201137788 contains more information to determine and/or distinguish between nearer and farther objects, which results in a wider range of data points (coarsely -grain) A more accurate map of the depth circle from 0 to 255. In some embodiments, the conversion to the custom depth range can be achieved by the equation: depth = GxB + R. In these embodiments, the maximum depth is 65, 280 = 255 x 255 + 255. Advantageously, by using this-conversion, the system and method of the present invention can easily compare custom depth values to raw pixel data. In some embodiments, the 1-key factor includes the color space (hue (H), saturation (S), and luminance (L)) e of the pixel, which can be determined by determining the level of saturation. / or the like and based on the decision assigning a depth value to the pixel to calculate three-dimensional information per pixel. For example, the ranges are for Η=0·360, Ι^()·1(Κ), (4), if the pixel has a high level of saturation or brightness (eg, s=6〇_1〇〇, l=3〇i〇〇), the depth generator 104 assigns a near-depth value (eg, a fan (if the depth range is set to 255 to 255)) to the pixel. Similarly, for example, if the pixel has a low level of saturation or brightness (eg, s = 〇 6 〇, B = 〇 3 〇) then the depth produces (4) 4 assignment - far depth value (eg, 5Q (if The depth range is set for 0 to 255)) to the pixel. In some embodiments, the key factor includes the amount of -value change in the pixel and the group of adjacent pixels. For example, the value change amount is one of the saturations between the pixel and the surrounding pixels. For example, based on the value: the variable 'depth generator 104 determines if the pixel is part of a plane (e.g., wall-desktop). Where depth generator 104 determines a portion of the pixel system plane, depth generator 104 assigns a far depth value to the pixel. ° J52792.doc -28- 201137788 In some embodiments, the key factor includes a position in each of the color segment frames =- or one of the plurality of objects. The position of the object can be two absolute positions (eg 'the object is located on the lower half of the color segment frame') or the position can be a relative position (eg, the object is in the color: segmented frame) The other - below the object). For example, referring to FIG. 2, the position of the color trowel & object 220 can be illustrated as being located in the lower half of the color segment frame 2, and the position of the color segment object 22 can be described as being located in the color point. Another object (not shown in the circle) in the segment frame 212D is squatted and/or the like. / 罙度器! 〇4 assigns a depth value for the pixel of the object based on the associated location data. For example, the depth generator 1 in the case where the position of the object is the lower position of the color segment frame (for example, under one color segment frame 2ud or below another object) 〇4 assigns a near depth value (for example, 2〇〇 (if the depth range is for 〇 to 255 and 疋)), σ » a parent pixel within the object (eg, each pixel of the color segment object). The position of the object is the upper position of the color segment frame (for example, the color segment object 218 is on the upper half of the color segment frame 2 丨 2C, or the color segment object 218 is attached to the In the case of a second object (not shown), the depth generator 1〇4 assigns a far depth value (eg, 5〇 (if the depth range is set for 0 to 255)) to each of the objects Pixel. Any range can be assigned to this depth value. In the case where the range is set to 255, the farthest depth value is set to 〇 and the most recent depth value is set to 255. In the case where the range is set to 65, 28 〇, the farthest depth value is set to 〇 and the most recent depth value is set to 6 5, 2 8 〇. 152792.doc -29- 201137788 For example, the position of the object is within a corresponding object of another color segment frame (for example, if the color segment frame 212C & 2i2D is superimposed, the color segmentation map The color segment object 22 in block 212D is superimposed on the color segment frame 212C and is greater than the color segment object 22 and extends around the color segment object 22〇. One color segment object (not shown) In the case of the above, the /wood generator 104 assigns a near depth value to each pixel within the object (i.e., the color segment object 220). For example, in the case where the position of the object is at an edge position of one of the color segment frames (eg, the color segment object 216 is at one edge portion of the color segment frame U2B), the depth is The generator 104 assigns a far depth value to each pixel within the object (eg, color segmentation object 216). In some embodiments, the key factor includes a ratio of the size of each of the one or more objects in each color segmentation frame to the size of one of the color segmentation frames. For example, the ratio can be calculated as the total pixel size of the pixel size of the color segment object 220 versus the color segmentation frame 212d. For example, if the pixel size of the color segment object 220 is 62, 5 pixels and the total pixel size of the color segment frame 2 (5) is 2, 〇73, _ pixels (ie, 1080x1920)' then the ratio Will be 62 5 〇〇 / 2 〇 73 6 〇〇 = 〇〇 3 〇. For each object, depth generator 104 may assign a depth value to each pixel within the object based on the ratio (eg, (in the case where the depth range is set to 255) depth value = 1 〇〇 (for a ratio between 〇〇1 and 〇1), 15() (for a ratio between ο" and G2), a fan (for a ratio between 0.2 and 0.5) or 255 (for a ratio between G5 and ))). 152792.doc 201137788

In some embodiments, the off M factor includes information indicating the location of one of the objects in the two-dimensional image and the location of one of the objects in the one-dimensional image. For example, + , , ° This location information can be used to determine the motion of objects between successive 2D images (you bite (for example, if a car is moving from left to right, compare two consecutive guides* (5) In the reverse, when the shirt is like, in the second continuous image (this is obtained at a later time, the shadows are φ, ...±, ?, d冢), & the car will be located on the right side.) Depth Generator 1 〇4 can determine that the position of the object in the two-dimensional image is different from the position of the corresponding object in the continuous two-dimensional image (for example, based on color segmentation of the color segment object in each of the two-dimensional images) The relative position of the corresponding color segment object in the frame, and assigning a near depth value (eg, 200 (in the case where the depth range is set to 255)) to the first two-dimensional frame Each pixel within a color segmented object. It will be appreciated that the various embodiments described above can be combined and need not be used separately (eg, the hue, saturation, and brightness of the pixel can be used in conjunction with the location of the color segmented object, etc.) Regarding step 316, the depth generator 1〇4 is based on The iso-basic gamut is calculated for the two-dimensional image (eg, 'by the depth map authoring unit 116. The depth map includes information indicative of three-dimensional information for each pixel in the two-dimensional image. As described above, the depth map is for The object within the two-dimensional image has increased depth information (ie, since the depth map includes a custom depth range for each pixel 'this increases the accuracy of the depth map). In some examples, the depth generator 1 When processing the two-dimensional image, the three-dimensional representation of one of the objects in the two-dimensional image (eg, via a stereo pair) may cause the object to enter the field of view and/or disappear within the field of view (eg, at the stand 152792.doc) • 31 · 201137788 When the left image remains the same, since the right image is drawn from one of the different points from the left image, the position of the object in the right image can enter the field of view and/or disappear in the field of view. The apparatus 1〇4 may need to adjust the depth map to account for such situations. Figure 5 illustrates an exemplary method 5 for manipulating an object in accordance with the present invention. At step 5〇2, The depth generator 104 selects a three-dimensional representation of one of the two-dimensional images (eg, X, a three-dimensional representation of the first object 208 of the two-dimensional image 202 of FIG. 2). At step 504, the depth generator 104 determines the two-dimensional representation. Whether the three-dimensional representation of one of the objects in the image causes the object to enter the field of view in a portion of the field of view in the two-dimensional image. For example, the depth generator 1〇4 can determine the two-dimensional image of the side of the first object 208. One of the invisible portions of 202 will enter the field of view while presenting the three-dimensional image (or for a stereo pair of images). If the depth generator 104 makes this determination, then the method 5 proceeds to step 506' Wherein the depth generator 1〇4 stretches the background behind the object, the stretch will enter the side of the object within the field of view, or a combination of the two to fill the portion of the object that enters the field of view. For example, in the case where the right side of the first object 208 enters the field of view, the depth generator ι 4 can stretch a portion of the right side of the first object 208 (and/or stretch the background 2〇4 and the view 206) A portion of the right side of the first article 208 is adjacent to fill in a gap that would otherwise be present at the right side of the first article 208. § If the measure generator 1 〇 4 does not make this determination at step 504, then the method 500 proceeds to step 508 and determines whether the three-dimensional representation of one of the two-dimensional images causes the object to be In a two-dimensional image, one part of the field of view enters the field of view. If the depth generator 104 makes this determination, 152792.doc • 32· 201137788 then the method 500 proceeds to step 51G and the depth generator i (m shrinks one side of the object to hide the portion of the object that disappears within the field of view For example, in the case where the left side of the first object 208 disappears within the field of view, the depth generator 104 may contract the portion to the left of the first object 2〇8 to compensate for the missing portion of the left side of the first object 208. In the case where the method does not make this determination at 508, the method proceeds to step 512, where the ten depth generator HM determines; t whether there are any remaining three-dimensional objects to be analyzed. In the case where there are remaining objects, the method Returning to step 5〇2 by selecting one of the remaining objects. Otherwise, the method proceeds to step 514 and terminates, π has analyzed and processed all three-dimensional objects (if necessary) (according to steps 504 to 5 1 When the depth generation II 1G4 calculates a depth map for the two-dimensional image based on the base gamut, the depth generator 104 may fine tune a previously generated depth map. The depth generator 104 stores (eg, via data)丨12) indicating data of a previously generated depth map. The previous graph is generated; the graph includes a depth map corresponding to each pixel in the two-dimensional image (eg, the depth map includes from 0 to each pixel) The data of the three-dimensional information of one of the depth ranges of 255. The depth generator HH may generate a new depth map based on the previously generated depth map, wherein the new depth map includes one of each pixel corresponding to a corresponding two-dimensional image. A wide range of two-dimensional information (eg, the range between 〇 and 65, 2 ribs as described above). Advantageously, the systems and methods described herein may be based solely on a previously generated depth map ( The previously generated depth map is quickly and efficiently fine-tuned, for example, via the equation: Depth-GxB+R, as described above. '152792.doc -33· 201137788 Figure 6 illustrates the image manipulation in accordance with the present invention An exemplary image 600. The two-dimensional image 602 includes a plurality of pixels 6〇4a, 6〇4b, 604C (respectively referred to as pixels 604). The pixel size of the two-dimensional image 6〇2 is the image along the two-dimensional image 602. One of the vertical sides (for example two The length of the pixel of the left side of the image 6〇2 is the product of the width of the image along the horizontal side of the two-dimensional image 6〇2—for example, the bottom side of the two-dimensional image 602. The two-dimensional image 6〇2 includes Object pixels 6 〇 6 八, 6 〇 6b, and 606 C representing one of the objects in the two-dimensional image 602 (collectively referred to as object pixels 606, the two-dimensional image 6 〇 2 includes boundary pixels 6 along the boundary of the object represented by the object pixel 606 〇 8 八, 6 〇 及 and 6 〇 8c (collectively referred to as boundary pixels 608). The reduced pixel image 61 〇 includes a plurality of pixels 612A, 612B (collectively referred to as pixels 612) 0 reduced pixel image 61 〇 includes expression minus The object pixels 6i4A, 6i4B (collectively referred to as object pixels 614) corresponding to one of the objects represented by the object pixel 606 in the small pixel image (4). The reduced pixel image 61 includes boundary pixels 616, 6i6B (collectively referred to as boundary pixels 616) along the boundary of the object represented by object pixel 614. Figure 7 illustrates an exemplary method 70 for autostereoscopic interpolation in accordance with the present invention (see Figure 丨, at step 7〇2, input unit 〇6 receives a first jersey image and a second two dimensional image The image, each of the two-dimensional images includes a pixel size. At step 7〇4, the processing unit 11 generates a reduced pixel image of each of the first two-dimensional image and the first two-dimensional image, wherein each A reduced pixel image includes less than one of the pixel sizes to reduce the pixel size. At step 706, the pre-processing unit 110 calculates boundary information for each of the first two-dimensional image and the second two, food V-images. At step 708, depth map 152792.doc -34 - 201137788 early 116 calculates a depth map for the first-reduced pixel image and the second reduction, wherein the depth map includes the indication of the first-decrease=image The image of the shirt and one of the second reduced pixel images or the information of the prime message. At the step, the depth map unit 116 is based on the boundary information of each of the ^^ image and the second two-dimensional image== a depth map of the prime image and the second reduced pixel image for 4 images The second two-dimensional image is formed by the base depth map. At step 712, the map generates a depth frame of the two-dimensional image and the second two-dimensional image, "ST is in the first Interpolating the second right image between the image and the second image includes a second profit point between the first-beneficial point of the first-two-dimensional image and the second two-dimensional image. Referring to step 702, the input unit 1〇6 receives a left image (a first two-dimensional shadow=) and a right image (a second two-dimensional image). The left image and the right image are the same field π (eg, object, landscape, etc.) Two different perspective views. The input unit 106 receives each two-dimensional image in the original pixel size. Referring to step 7〇4, the pre-processing tl 11 0 generates a reduced pixel for each of the left image and the right image. The image-per-reduced pixel image includes less than the left image and the right "image size of the original pixel size. The pixel size is reduced. For example, if the image is not shown in Figure 6, the pre-processing unit U is based on the two-dimensional image 6 〇2 produces a reduced pixel image 610 ° reduced pixel image 610 has fewer pixels For example, the arrows 650A, 65GB (collectively referred to as arrows 65G) overlaid on the 3D image 6〇2 divide the dimension image 6() 2 into squares including four pixels. The preprocessing unit m translates the 2D image 6〇2 152792.doc • 35- 201137788 into a reduced pixel image 61〇(1) set to the four image sympleries: the group of reduced pixel images (4) pixels in the 2D image 602 corresponds to the main square, and the main pixel. For example, by the edge of the pixel, therefore, three of the four pixels of the =7 order are translated into a test, 丨^3 edge pixel 6〇8B two-dimensional The image of the image 602 is imaged by the edge image 616 Β β I, the image 602 and the reduced image, and the pixel shirt image 61 is only executed by the exemplary translation technique (Example 1 turn: Small pixel images can be mapped using different pixel mappings. The more/less scenes/silly in the quasi-images, the material ^ " Decrease multiple pixels in the pixel image, =). In addition, the reduced pixel size produced by the pre-processing unit 11 can be adjusted (e.g., (10) pixels, (10) pixels x 200 pixels, etc.). Since each image can have pixels that other images may not have (for example, due to the viewpoint of the image), each of the left and right images can be reduced. For example, as described above, only one side of one of the left and right images may be visible from an angle, there may be - hidden points &/or the like. Referring to step 706, the processing unit 4 calculates the boundary element | for each of the left image and the right image. During this step, it is more likely that the points in the left and right images (eg, one or more points) are more likely to be used to ultimately produce a depth map between the left and right images. (eg, at steps 708 and 710) important points (ie, object boundaries). The pre-processing unit Π0 calculates edge information for the left and right images, and uses edge information to determine one of the stereo images. A common point is provided for comparison 152792.doc •36· 201137788 One of the left and right image reference points. For example, the pre-processing unit "〇 can determine that the common point between the left image and the right image is the pixel located on the side of the object in the image (example %, pixel 608C of two-dimensional imaging). Advantageously, by finding a common point between the left image and the right image, the preprocessing unit 11G does not need to match multiple points in the left and right images. In some embodiments 'the edge points are selected based on the color segmentation frames generated from the left and right images (rather than the original left and right images) (eg, the color segmentation unit 114 is for color as described above) Each color used in the segmentation process produces a plurality of color segmentation frames for the left and right images). Therefore, the color segmentation images can be used to determine object boundaries. For each pixel, edge generation unit 118 determines whether the pixel is a boundary point and produces a set of pixels including boundary information (e.g., boundary pixels 6〇8). Advantageously, the system and method of the present invention achieves a significant time savings over the generation of boundary information by other means by performing color segmentation and identifying boundaries based on the color segmentation frames. Referring to step 708, the depth map unit u 6 calculates a depth map for the first reduced pixel image and the first reduced pixel image. As described with reference to Figure 8 below, depth map unit 116 may calculate information regarding the difference in pixel locations in the reduced pixel image as the depth map is generated. Referring to step 7 〇, based on the boundary information of each of the two left and right images (the boundary information generated at step 7〇6) and the depth map of the reduced pixel images (at step 708) The depth map generated is used to calculate the depth map of the left and right images. For example, in some embodiments, based on pixel values corresponding to the left reduced pixel image 152792.doc -37- 201137788 and the right reduced pixel image (eg, based on the boundary between the reduced pixel and the image (4) The depth value of the pixel 616 is used to calculate the depth value of each side #pixel^, such as the boundary pixel 608 of the two-dimensional image 602. In a practical example, the average value and the value of the depth map of the reduced pixel image are determined (for example, based on the corresponding boundary pixel of the left-reduced pixel image and the right-reduced pixel image). The data is used to determine the depth information of the depth circle of the remaining pixels (ie, pixels that are not border pixels). Referring to steps 708 and 71, the depth generator 1〇4 (i.e., depth map authoring unit 116) can identify pixels of an object that is visible in __ images but not visible in other artifacts. For example, in the case where the object in the left image contains a pixel on the outermost left side of the object, the pixel may not be visible in the right image. This is because the outermost pixel on the left side of the object is moved to the field of view (4) when the object is viewed from the vantage point of the right image due to the favorable angular difference between the two images. Similarly, one of the pixels on the right side of the object that is not visible when viewing the object from the vantage point of the left image can enter the field of view in the left image (i.e., 'becomes visible'). Therefore, the conversion unit 108 can generate (or interpolate) the third two-dimensional image based on the depth maps of the left two-dimensional image and the right two-dimensional image, and the hidden pixels identified by the third two-dimensional image include one of entering the field of view. A region (that is, a pixel that is not visible in the left image but visible in the right image) or a region that is in the field of view (that is, a pixel that is visible in the left image but not visible in the right image) . ...~Step 712 'Conversion unit 1() 8 interpolates between the left image and the right image. For illustrative purposes, assume a 〇. One of the reference angles obtains the left image 'from 2' at a favorable point of 152792.doc -38- 201137788. One of the vantage points to obtain the right image (actually, the angle of any suitable range can be used). Therefore, by generating a depth map of the left two-dimensional image and the right two-dimensional image, the conversion unit 1 8 can now be between the advantageous points of the left image and the right image (for example, between Q and 2) Favorable points produce - images. For example, the 'conversion unit ι 8 can generate a third two-dimensional image based on the depth maps of the left two-dimensional image and the right two-dimensional image having a favorable point of 0·5 〇_53, G.l, and the like. 8 illustrates an exemplary method 800 of assigning depth information in accordance with the present invention, wherein the depth is calculated based on the difference in pixel locations (eg, the difference in pixel locations of each of the objects t in the reduced pixel image). The figure (for example, the left-reduced pixel image and the depth map of the right-reduced pixel image). At step 802, the depth generator 1〇4 compares the two corresponding pixels (eg, via the depth map authoring unit U6), wherein the first pixel is located within the first-reduced pixel image (eg, a reduced pixel image of the left image, and one of the two corresponding pixels is located within the first reduced pixel image (eg, a reduced image of the right image (four) image). At step 804, depth generator 104 determines that the two pixels are missing the parent fork. If the pixels are separated (for example, when the left image and the right image are superimposed, the corresponding pixel of the left image is located on the right side, and the right pixel is located at the right side of the pixel from the left image - the distance is 1) Go to step 806' where the depth generator 104 calculates how far apart the first pixel of the image is from the second pixel of the right image. At step 808, the depth generator 1〇4 assigns to indicate two of the depth maps. The depth information of the pixel is information indicating that the object is away from one of the three-dimensional views of the object containing 152792.doc -39 - 201137788. If at step 804 the depth generator 1 〇 4 determines that the pixels are intersecting (eg When the left image and the right image are superimposed, the corresponding pixel of the left image is located on the right side, and the corresponding pixel of the right image is located at a distance from the left side of the pixel of the left image. The method 800 proceeds to step 81 and the depth generator calculates One of the degrees indicating that the left pixel intersects the right pixel (eg, the distance between the left pixel and the right pixel) is a value. At step 812, the depth generator 1〇4 assigns an indication of the depth. Two of the corresponding depth information of the pixel to indicate that the object is close to one of the three-dimensional views of the object. Steps 808 and 812 proceed to step 814, wherein depth generator 104 determines if any The remaining pixels that have not yet been calculated (eg, the left-reduced pixel image and any pixels of the right-reduced pixel image) ^ If there are remaining pixels, the method 800 proceeds back to step 8〇2. Otherwise, the method proceeds to steps 816 and depth generator 104 completes computing depth map 816. Figure 9 illustrates an exemplary method 900 for manipulating objects in accordance with the present invention. At step 902, depth generator 104 (i.e., color segmentation unit 114 and / Or edge generation unit u 8) generates a plurality of color segmentation frames based on the two-dimensional image, wherein each color segmentation frame includes one or more objects. Depth generator 104 can be configured to generate any number of colors Segmentation frame. For example, 深度 'Depth generator 104 can be configured to generate 1 色彩 color segmentation frame, each color segmentation frame corresponds to a different color (example) 'Dark red, light red, etc.." At step 904, depth generator 104 selects a color segmentation frame from the plurality of color segmentation frames. At step 906, the depth generator verifies the color segment. The segment frame contains one or more 152792.doc • 40· 201137788 :: objects, each of which includes an identifiable boundary line. If at step 906 the depth generator 104 determines that the color segmentation frame does not contain - two = The polymer member's method proceeds to step _ and the depth generator 1〇4 discards the color segmentation frame (ie, the frame is not used to calculate a depth if the depth generator (10) determines the color segment at step _ The frame comprises one or more internal polymer parts, for each of the pixel images of the first and second pixels, and the pixel image of the second reduced pixel image. The device 104 sets a boundary point of the pixel based on the color of the segment to be inconsistent, wherein the boundary point includes the data of the (four)-boundary point. Following step 906, each color segmentation frame should be closed to make a deep slap. In general, this means that the color segment frame should contain a bounded "built-in map. There are multiple objects in the color segment frame (for example, if the original image is "human * ~" like humans, And the specific color segmentation package 3 represents a separate part of the human body, such as the hand, the foot = the single object 'and one of the objects of the face 'but the other part of the human body is not displayed for the particular segment frame) In the case where the object of step 9〇6 has a defined boundary, the color segmentation frame is still considered to be closed. The one of the color segmentation frames is not closed. One of the pixels is composed of a scatter or scatter and there is no frame of definable subject of interest within the segment of the color. In some embodiments, the depth generator 1 可 4 can contain and control the three-dimensional object u generates (eg, when generated by the conversion unit (10)) a depth map associated with the configurable parameter. For example, the configurable parameter can be set or adjusted (eg, via "λ „ _ 雨早 70 106 ) to configure the conversion unit 108 to such 152792 .doc 41 201137788 The shirt image is presented as a two-dimensional image, a one-dimensional image of the r-dimensional image, formatted for display on a cellular phone, etc. Class gentleman ^ ^ cheek-like, In some instances, the configurable parameter can be adjusted to control the amount of parallax of the three-dimensional image to maintain the correct size of the objects in the three-dimensional scene to adjust the three-dimensional material for various screen sizes and spatial sizes (eg, not only for ―The size of a specific space is calculated once, and it is smashed and simplistic. Thousands of images are adjusted for use in various display applications during operation. ^ 某 彳 彳 丨 ^ ^ ^ ^ ^ ^ ^ ( ( ( For example, image security to set or adjust the configurable parameter. As illustrated in Figure 10, the system between the virtual distance and the real distance according to the present invention - the example of the formula 1-1 straight axis shows the virtual distance (10) 2 ( The example 'measured in meters (10)), and the horizontal pumping shows the true distance view (for example, measured in meters (m)). The belt 6 is for one of the first objects (for example, one person) Thickness and true distance 1 One of the virtual distances (7) compared to 〇4 indicates that the band 1008 is represented by one of the virtual distances 1 〇〇 2 of one of the second objects (eg, a tree) compared to the true distance 1004. 1 〇展不,, the relationship between the virtual distance of the first object and the second object and the real distance 1004 is a 1:1 relationship, which is correctly maintained when the object in the two-dimensional image is rendered as a three-dimensional drawing. Distance and Thickness. Figure 11 illustrates a consistent exemplary graph 00 of the relationship between virtual distance and true distance in accordance with the present invention. The vertical axis shows the virtual distance η and the horizontal axis shows the true distance 1104, as described with reference to Figure 10 The band 1106 is for the virtual distance 1102 of the thickness of the first object compared to the true distance 1104. The band 1108 is represented by one of the virtual distances 1102 for the thickness of the second object compared to the true distance 丨〇4. Line 111〇 shows what is the ratio of the virtual distance 1102 to the true distance 1104 in the case of being properly maintained 152792.doc *42- 201137788. However, curve 1112 shows how the ratio of the first object to the second object depends on the camera parameters and the screen size (eg 'the camera is narrow, the field of view is wide, or the screen size is small: can occur "Cardboard effect"). As shown by curve 1112 and strip 11〇8, the virtual thickness of ιι〇8 is thinner than the true thickness, which results in a "cardboard effect" (for example, the tree looks thinner than it actually looks in real life). Similar to Fig. 11, Fig. 12 illustrates an example (four) phantom relationship between a virtual distance and a true distance in accordance with the present invention. The straight axis shows the virtual distance 1202, and the horizontal axis shows the true distance 12〇4, as shown in Fig. 1 and Fig. 11. Line 1206 is representative of one of the virtual distances 1202 of the first object compared to the true distance 12〇4. Line 1208 is represented for one of the virtual distances 1202 of the second object compared to the true distance 12〇4. Line 121 〇 shows one of the relationships between virtual distance 1202 and true distance 丨 204 when properly maintained. What is the ratio of ι:ι. Similar to curve 1112 of Figure u, curve 1212 shows how the ratio of the first object to the second object is dependent on camera parameters and screen size (eg, narrower camera angles, narrower camera angles, or screen size) When the system is large, a midget effect can occur. As shown by curve 1212 and line 1206, the virtual distance 12〇2 is shorter than the true distance 12〇4, which results in one effect commonly referred to as the “miniaturization effect” (for example, people seem to look at their real life more than they are in real life) It's short). Advantageously, the systems and methods can be configured to take into account the board effect and miniaturization effects as described above when determining the depth information of the depth map. For example, the depth information is configured to maintain a 1:1 ratio by multiplying the calibration curve function by I52792.doc -43· 201137788. In some embodiments, for stereo imaging, a distorted image can be used for both the left image and the right (five) image. For example, the 'original image' can be used as the "center point" of the viewpoint when converting a single two-dimensional image into a stereo pair. Therefore, when a stereo pair is generated for the original image, the left image of the stereo pair is presented to the left of the favorable point of the original image, and the stereo image is presented at a favorable point on the right side of the advantageous point of the original image. Right image. In this embodiment, the original image used to generate the stereo pair is not used as the left or right image of the pair. The above systems and methods can be implemented in digital electronic circuits, computer hardware, hardware, and/or software. The embodiment may be implemented as a computer program product (ie, tangibly embodied in a computer program in an information carrier, and the implementation may be in a machine readable storage device 1 executed or controlled by the data processing device The operation of the data processing device. For example, the actual solution can be a programmable processor, a computer and/or a plurality of computers. ^ The computer program can be in the form of a programming language (including compiling and / writing) 'and may be in any form (included as a stand-alone program, as a secondary routine, component and/or other computer program suitable for use in a computing environment. A computer program may be deployed to be on an electric island or in a Executing on a plurality of computers at a location. The steps may be performed by executing a computer program to perform the functions of the present invention by performing an operation on the input data - or a plurality of programmable portions: execution. The method steps may also be performed by dedicated logic The circuit is executed, and a device is implemented as a dedicated logic circuit. For example, the circuit can be - 152792.doc *44-201137788 FPGA (field programmable gate array) and / or an ASIC (Application-Specific Integrated Circuit). Modules, sub-normal and software agents may refer to computer programs, processors, specific circuits, software and/or hardware that implement the functions. A processor suitable for executing a computer program includes both a general purpose microprocessor and a dedicated microprocessor, and any one or more processors of any kind of digital computer. In general, a processor from a read only memory The body or a random access memory or both receive instructions and data. A basic component of a computer is a processor for executing instructions and a device for storing instructions and data or a plurality of devices. A computer may include a storage device (eg, a magnetic disk, a magneto-optical disk, or a compact disk) for storing data, or operatively coupled to receive data from the mass storage device and/or transmit data to the device. Such mass storage devices. Data transmission and instructions may also appear on a communication network. The information carrier suitable for embodying computer program instructions and data contains all forms of non-volatile memory. And including, by way of example, a semiconductor memory device. Such information carriers may, for example, be EpR〇M, EEpR〇M, flash memory device disk, internal hard disk, removable disk, magnetic A disc, CD-ROM and/or DVD_R.M. The processor and memory may be supplemented by dedicated logic circuitry and/or incorporated into dedicated logic circuitry. To provide for interaction with the user, the techniques described above may be implemented with For example, the display device can be a polar ray tube (CRT) and/or a liquid crystal display (LCD) monitor. Interaction with a ^^ can be (for example) One of the user's information displays and enables the hunter to provide input to the computer (for example, interacting with the system interface) 152792.doc •45· 201137788 One of the keyboards and an indicator device (for example, - a mouse or a Trackball). Its # type of device can be used to provide

(for example) providing feedback to the user of the user in any form of sensation feedback (eg, visual feedback, feedback, or tactile feedback). Inputs from the user can be received, for example, in any form or pattern, including audible, tactile inputs. ^

The above techniques can be implemented in a distributed computing system including a backend component. The backend component can be, for example, a data server, an intermediate component, and/or an application server. The above (4) one of the solid H front-end components can be used in a decentralized system. The front end component can, for example, be a client computer having a graphical user interface, a user interface with which the user can interact with the exemplary embodiment, and/or Other graphical user interfaces of a transmission device. The components of the computing system can be interconnected by any number of data communication forms or media (e.g., a communication network). Examples of communication networks include a local area network (LAN), a wide area network (WAN), the Internet, a wired network, and/or a wireless network. The system can include a client and a server. A client and a server are generally remote from each other and typically interact through a communication network. The relationship between the client and the server is generated by a computer program running on a separate computer and having a client-server relationship with each other. A packet-based network may include, for example, the Internet, a carrier Internet Protocol (IP) network (eg, regional network (LAN), wide area network (WAN), campus area network (CAN) ), Metro Network (MAN), Family Area 152792.doc • 46· 201137788 Domain Network (HAN)), _Takada ττ »

Private IP network, an IP private switch (IPB - wireless network (such as I + ... line access network (RAN), 802.1 1 network, • 6, road, general packet radio service (GPRS) network, H]perLAN), and/or other packet-based networks. Circuit-based networks can be (3 (for example) public switched telephone network (PSTN), private branch exchange (PBX), wireless network ( For example, ran, Bluetooth, code division multiple access (CDMA) networks, time division multiple access (tdma) networks, global mobile communications (GSM) networks, and/or other circuit-based networks. The transmission device may comprise, for example, a computer, a computer with a viewer, a telephone, an IP telephone, a mobile device (eg a cellular telephone, a personal digital assistant (PDA) device, a laptop Computer, email device, and/or other communication device. The browser device includes, for example, a global information network device (eg, available from Microsoft Corporation)

Microsoft® Internet Εχρ1〇_, one of the computers available from M〇ziiia, _ Firefox) (example #, desktop, laptop) The mobile computing device contains, for example, a number of people Assistant (PDA). Each of the included, including, and/or plural forms are open-ended and include the recited components and may include additional components not listed, and/or are open-ended and include the recited components Combinations of one or more of the recited components. Those skilled in the art will recognize that the present invention may be embodied in other specific forms without departing from the spirit or essential characteristics of the invention. The following is intended to be illustrative, and not to limit the invention of the present invention as described in the 152 792- doc s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s All changes in the meaning and scope of the equivalent inner valley belonging to the scope of the patent application. [Simple description of the schema] An exemplary 3D rendering system; Color segmentation image 1 derived from an image illustrates the root Figure 2 of the present invention illustrates an exemplary diagram in accordance with the present imagery; Base Figure 3 illustrates an exemplary method for a color-coded color domain in accordance with the present invention; Segmented Image Pulling Figure 4 illustrates An exemplary method for defining an edge of an object within an image in accordance with the present invention; Figure 5 illustrates an exemplary method for manipulating an object in accordance with the present invention; 'an example of a stereoscopic interpolation of an example image Example of Depth Information FIG. 6 is a diagram showing an image manipulation method according to the present invention; FIG. 7 illustrates an automatic method according to the present invention; FIG. 8 illustrates a method for assigning according to the present invention; An exemplary method for manipulating an object of the present invention; FIG. 10 illustrates an exemplary diagram of a relationship between a virtual distance and a true distance gate in accordance with the present invention; FIG. 11 illustrates a virtual distance to a true distance in accordance with the present invention. 152792.doc -48- 201137788 one of the relationship diagrams; and FIG. 12 illustrates the relationship between the virtual distance and the true distance in accordance with the present invention An example schema of the relationship. [Main component symbol description] 100 3 D reproduction system 102 computer 104 depth generator 106 input unit 108 conversion unit 110 pre-processing unit 112 database 114 color segmentation unit 116 depth map creation unit 118 edge Generation unit 120 3D experience unit 122 Input device 124 Data 126 Display 202 2D image 204 Background 206 Foreground 208 First object 210 Second object 212 Color segmentation frame 152792.doc -49- 201137788 214 Color segmentation object 216 Color division Segment object 218 color segment object 220 color segment object 602 2D image 604 pixel 606 object pixel 608 boundary pixel 610 reduced pixel image 612 pixel 614 object pixel 616 boundary pixel 650 arrow 1002 virtual distance 1004 true distance 1006 with 1008 band 1010 Line 1102 Virtual Distance 1104 Real Distance 1106 with 1108 with 1110 Line 1112 Curve • 50- 152792.doc 201137788 1202 Virtual Distance 1204 Real Distance 1206 Line 1208 Line 1210 Line 1212 Curve I52792.doc -51

Claims (1)

201137788 VII. Patent application scope: 1. A computerized method for extracting a primary color gamut from a plurality of color segmentation images. The method comprises: storing data of a non-two-dimensional image in a data storage device, The two-dimensional image includes a plurality of pixels; a computer-color segmentation unit generates a plurality of color shirt segment frames based on the two-dimensional image, wherein each color segment frame includes one or more objects; Segmenting I elements for each of the color segmentation frames to generate a primary color gamut based on the one or more objects; and wherein the computer t-depth map element is based on the primary color gamut for the two-dimensional image A calculation-depth map, wherein the depth map includes data indicative of three-dimensional information for each pixel of the two-dimensional image. 2. The method of claim 1, wherein the step further comprises allowing adjustment of the Bezier point of the Bezier graph of one of the objects. 3. The method of claim 2, further comprising: defining, for each color segment frame, an edge of each of the one or more objects, including: 〃 automatically calculating one of the objects图图兹固°海贝炫图 includes a plurality of Bezier points for the piece and a plurality of Bezier curves connecting the plurality of Bez points; 'receiving instructions to adjust the information of the Bates chart; and generating based on the data A more detailed view of one of the objects, wherein the more detailed map includes a plurality of additional points combined with the plurality of Bezier points. 152792.doc 201137788 4. The method of claimants, which produces the color segmentation frames to define a plurality of basic colors, each of which is a segment of the basic color frame. The method of claim 4, wherein each of the plurality of basic colors from the plurality of basic colors includes a range of color values, including a color value within the range of the predefined color values. The system is associated with generating a color segmentation frame based on the basic color. 6. The method of claim 5, comprising: receiving data indicative of a range of color values of one or more of the plurality of basic colors; and wherein the data is for each of the one or more colors Adjust the range of pre-existing color values. 7. The method of claim 4, wherein the plurality of basic colors after the refilling of the gold f, the color of each of the plurality of basic colors comprises a color Baton pair, the color pair comprising a light color and a deep color . 8. As in the method of claim 4, 苴中命益廿, 八^ is a number of basic colors including brown and beige. 9. The party of claimants, wherein calculating the depth map comprises: determining that the two-dimensional representation of the object in the two-dimensional image causes a portion of the object that is not in the field of view to enter the field of view in the two-dimensional image. And stretching one of the back of the object, the back of the tooth, one side of the object, or both to fill the portion of the object that enters the field of view. 10. Calculating the depth map in the method of claim 1 includes determining that the two-dimensional representation of the object in the two-dimensional image results in the field of view in the two-dimensional I52792.doc • 2. 201137788 image. The portion of the object disappears within the field of view; and the side of the D-object is shrunk to hide that portion of the object that disappears within the field of view. 11. The method of claim 1, wherein calculating the depth map comprises calculating the information for each pixel based on a one-key factor of the plurality of color segment frames. 1. The method of claim η, wherein the key factor includes the hue, saturation, and brightness of the pixel, and the calculation includes: determining the level of saturation, the level of brightness, or both—high is a low level; And assigning a near depth value to the pixel if the level of saturation or brightness is a high level; or the depth value is given to the low level when the level of saturation or brightness is a low level Pixel. m: the method of member 11 wherein the key factor includes the amount of -value change in the pixel and the 4-pixel group, and the calculating comprises: determining whether the pixel is a portion of a plane based on the value change; and in the pixel system In the case of a portion of the plane, the pixel is assigned. The deciduous value is as in the method of claim 11 wherein each of the one or more objects in the key factor frame is in a positional color including for each object: and the (four) object ^ position is Color segmentation frame - In the case, the assignment-near depth value is given to each _pixel in the object 152792.doc 201137788 2: This position is the color segment frame - the upper position of the month / Assigning a far depth value to each pixel in the object, assigning a near depth value to the bar in the other color component at the position of the object, and the corresponding month of the segment The position of the object · 忐 n n &lt; mother - like the official object is in the color segment frame 15. 16. 17. In the case, 'assign a far depth value to each edge of the object The method of item U, wherein the key factor comprises a size of each of the one or more objects in each of the color segment maps, the size of the color: A depth value is assigned to each of the objects based on the ratio for each item. = method of not item U, wherein the key factor includes information of one of the objects in the image w^-dimensional shirt (four) (bit m continues to one of the positions of the two-dimensional image, and the calculation includes: the corresponding object determines the two-dimensional The location of the object in the image is different from the location of the corresponding object in the dimension image; and successively assigning a near depth value to each pixel in the object. The method of claim 1 further comprising: In the data storage device that will indicate the previously generated depth map, the depth map includes three-dimensional data of the dry (four) jitter; and each pixel of the H-turn image is generated based on the previously generated depth map - a new depth The depth map includes information indicating the range of information per pixel corresponding to the two-dimensional image. #八二152792.doc 201137788 18·Method of clearing item 1 wherein calculating the depth map includes applying one or more bodies The rules are adjusted to adjust the depth map 'each rule is configured to adjust the depth map based on human perception of one of the one or more objects in the one-dimensional image. A system for priming a color gamut, the system comprising: a cultivating storage device configured to store a resource indicative of a two-dimensional image comprising a plurality of pixels; a color segmentation unit and the data The storage device communicates, the color segmentation unit configured to: generate a plurality of color segmentation frames based on the two-dimensional image, wherein each color segmentation frame includes one or more objects; and for the color segmentation maps Each of the frames generates a primary color gamut based on the one or more objects; and a depth map unit that communicates with the color segmentation unit and the data storage device, the depth® unit being configured to be based on the f The base color gamut calculates a depth map for the -dimensional image, wherein the depth map includes data indicative of three-dimensional information for each pixel of the two-dimensional image. '2〇· As in the system of claim 19, the further unit is configured to The step includes an input unit, the input material is stored in the data receiving the data indicating the two-dimensional image and is stored in the data storage device; and the production receiving indication is generated by a previously generated depth map. The data of the map is stored in the data and will be indicated in the prior data storage device. 152792.doc 201137788 21. 22. 23. 24. The system of claim 19, further comprising configured to target the two dimensional Each object in the image produces an edge generation unit of edge information. The system of claim 19, further comprising a three-dimensional experience unit configured to apply - or a plurality of experience rules to adjust the depth map, each ruled Configuring to adjust the depth map based on one of the one or more objects in the two-dimensional image. A computer program product 'is tangibly embodied in a computer readable storage medium' The operation of causing a data processing device to execute the following steps: storing the data indicating the two-dimensional image in a data storage device, the two-dimensional image comprising a plurality of pixels; and the color segmentation unit of the computer is based on the two-dimensional The image generates a plurality of color segmentation frames, wherein each color segmentation frame includes one or more objects; Each of the color segmentation frames generates a primary color gamut based on the one or more objects; and wherein the computer-depth map unit calculates a depth-depth® for the two images based on the primary color gamut, wherein the depth map Includes information indicative of three-dimensional information for each pixel of the two-dimensional image. A thunder brain for autostereoscopic interpolation, the soil is received by an input unit of a computer - a second-dimensional image, each two-dimensional image comprising a pixel size; and the pre-processing unit of the computer is directed to the first Each of the two-dimensional image and the one-dimensional image produces a reduced pixel image, wherein the <electric chesting method includes: the image and a 152792.doc 201137788 a reduced pixel image includes less than the a pixel size in which one of the pixel sizes is reduced; calculating, by the preprocessing unit, boundary information for each of the first two-dimensional image and the second two-dimensional axe image t; Calculating a depth map of the reduced pixel image and the second reduced pixel image, wherein the depth map includes one or more objects indicating the first reduced pixel image and the second reduced pixel image The data of the three-dimensional information; and the depth map unit of the computer is based on the boundary information of each of the first two-dimensional image and the second two-dimensional image and the first-reduced pixel image and the first ^ The reduced degree image of the pixel image (4) The first two-dimensional image and the second two-dimensional image calculate a depth map. 25. The method of claim 24, wherein the step of generating a third two-dimensional image based on the depth image of the second-dimensional image and the second two-dimensional image, wherein the third two-dimensional image comprises An advantageous point between the first advantageous point of the first two-dimensional image and the second advantageous point of the second two-dimensional image. Uniformly ~ 26. As in the method of claim 24, the calculation of the depth in the m* pixel image and the pixel image of the pixel is calculated based on the reduction: in the object in the prime image The pixel location of each of them produces the material. The method of claim 26, further comprising: comparing two corresponding images, wherein the pixel is located in a first pixel image of the first subtraction/two corresponding pixels, and the two corresponding images 152792 .doc 201137788 One of the first pixels is located in the pixel image of the _,, &amp; __ music-reduced pixel; the calculation refers to the first pixel and the second value of the TF; and how far the pixel is open The distance-assignment indicates the depth information of the two corresponding pixels in the depth map to indicate that the object is away from the viewer. The method of claim 26, further comprising: the two corresponding pixels, wherein one of the two corresponding pixels is located in the first-reduced pixel Shadow (4), and the second pixel system of the two corresponding pixels is located in the second reduced pixel image; the calculation refers to the first pixel and the first value; and the pixel parent degree is one of the cross assignments indicating The data of the depth information of the two corresponding pixels in the depth map is used to indicate that the object is close to one of the three-dimensional views containing the object. 29. The method of claim 24, wherein the constant-size leaf is calculated as the product of the pixel length and the pixel width of each corresponding two-dimensional image. The method of claim 24, wherein calculating a depth map for the one-dimensional image and the second two-dimensional image comprises: calculating, for the boundary information based on the first-two-dimensional image and the second two-dimensional image a depth map of a boundary pixel; and teaching the depth information of the depth map of the suspension by the data of the corresponding boundary pixel adjacent to the first-reduced pixel image and the first image. The method of claim 24, wherein calculating the boundary information comprises: generating a plurality of color segmentation frames based on the two-dimensional image, wherein each color segmentation frame comprises one or more objects; Setting a boundary point indicator of the pixel based on the color pixel of the first reduced pixel image and the second reduced pixel image, wherein the boundary point indicator includes the indication Whether the pixel is a boundary point data. 32. The method of claim 3, comprising verifying that each of the plurality of color segment frames comprises one or more inner polymer members, each inner polymer member comprising an identifiable boundary line. 33. The method of claim 31, wherein calculating the depth map comprises identifying a hidden pixel of the first reduced pixel image by identifying a visible pixel in the first reduced pixel image, wherein the visible pixel There is no corresponding pixel in the second reduced pixel image. 34. The method of claim 33, further comprising generating a third two-dimensional image based on the depth image of the first two-dimensional image and the second two-dimensional image, wherein the second two-dimensional image comprises the first An advantageous point between a first vantage point of one of the two-dimensional images and a second vantage point of the second two-dimensional image, and wherein the third two-dimensional image includes one of entering the field of view based on the identified hidden pixel Zone or area that disappears within the field of view. 35. - System for (four) dynamic stereo interpolation 'The system includes: a round-in unit configured to receive a first two-dimensional image and a second two-dimensional image, each two-dimensional image including - pixel size a pre-processing unit that is in communication with the input unit, the pre-processing unit 152792.doc 201137788 configured to: generate a reduced-pixel image of the first two-dimensional image and the second two-dimensional image, each of which a pixel size smaller than a pixel size of a pixel image package, and a pixel size for each of the second-dimensional image and the first calculated two-dimensional image and the first calculated image And the pre-processing element is configured to: "early 7", "the depth map is for the first reduction of the silly sound search "De β * &amp; like calculating a deep Lu Tu loading by ... δ Λ second reduction The image circle of the pixel image includes data indicating the second-dimensional poorness of the first-decrement=prime: image and the second reduced pixel image, and/or a plurality of objects, and: Each of the two-dimensional image and the second two-dimensional image = boundary information and the first reduced pixel image And the depth map of the second reduced image is used to calculate a depth map for the first two-dimensional image and the second one-dimensional image. 36. 37. The system of the long item 35 includes The pre-processing unit communication) knife &amp;single&apos;2 color segmentation unit is configured to generate a plurality of color segmentation frames based on the two: the image segmentation frame I includes one or more objects. The system of item 5, wherein the step further comprises configuring the unit to convert the unit based on the depth/dimensional image, wherein the third two-dimensional scene "image comprises a first advantageous point between the first two-dimensional image An advantage between the second advantageous point of the second-dimensional image. 152792.doc 201137788 recognizes a computer program that produces π〇σ, which is tangibly embodied in a computer readable storage medium, the computer program product comprising instructions operable to cause a data processing device to perform the following steps: receiving a first two a dimensional image and a second two-dimensional image, each of the two-dimensional images including a pixel size; for each of the first two-dimensional image and the second two-dimensional image, a reduced pixel image is generated, wherein each- The reduced pixel image includes a pixel size that is smaller than one of the pixel sizes; for each of the first two-dimensional image and the second two-dimensional image; ° for the first-reduced pixel image And calculating, by the second reduced pixel image, a depth map, wherein the depth map includes data indicating three-dimensional information of the first-reduced pixel image and the one or more objects in the second reduced pixel image; And the depth map based on the boundary information of each of the first two-dimensional image and the second two-dimensional image and the first-reduced image (four) image and the reduced-pixel image The first two-dimensional image and the second two-dimensional image calculate a depth map. 152792.doc -11 ·