US 20050062742 A1
3-D views of a 2-D image containing objects are provided by simulation of inter-object volume and intra-object volume. Intra-object volume is simulated by defining a plurality of volumetric slices from the object and assigning relative volume values to each. Multiple views of the object are simulated by displacing each slice in accordance with the defined volumetric effect as part of an animation sequence simulating progressive movement about a focus point of the object.
1. A method for producing an input image for application from a two dimensional image onto an image surface adapted to present interlaced frames of the input image, comprising:
receiving an input image, the image including at least one object;
receiving layer definitions for a plurality of layers for at least one object of the input image, the layer definitions include data for object layers that are copied from the object by defining an area containing coplanar portions of the object;
receiving an indication of the total displacement for the object animation;
determining a relative displacement for each layer from the plurality of layers;
generating a sequence of image frames by progressively offsetting said plurality of layers in accordance with relative layer displacement and a total animation displacement; and
interlacing the animated sequence of frames to a combined image to provide an input interlaced image for application to said image surface.
2. The method of
3. The method of
4. The method of
5. The method of
6. The method of
7. The method of
8. The method of
selecting a focus point, the focus point associated with a layer from the plurality of layers which remains static when an observation point for the animation is varied;
assigning a relative displacement for each layer in the plurality of layers by reference to the received total displacement for the object, the number of frames in the frame sequence, and the number of layers between the layer and the focus point;
setting the layer displacement to a first extreme orientation by displacing each layer in the reverse to the relative displacement direction and by multiplying the reverse relative displacement by half of the total frame number; and
displacing each layer in accordance with the assigned relative displacement and frame number in the sequence, the displacement of each layer in each frame provided by multiplying the frame number by the relative displacement to arrive at the layer displacement at a second extreme orientation to provide a symmetrical sequence of animation frames for the object.
9. The Method of
This application is a continuation of and claims priority from U.S. patent application Ser. No. 10/211,500, entitled “Method and System for 3-D Object Modeling” filed Aug. 2, 2002, now pending, which is incorporated herein by reference.
The present invention relates to graphical modeling, specifically the invention relates to volumetric simulation from a two dimensional input image.
Stereoscopic systems present a viewer with an image of an object such that the object appears to have depth. The wide variety of techniques to provide such images commonly involve presenting different views of the object to each eye to simulate such images. Traditional stereoscopic systems employ an apparatus that keeps the left and right eye images directed solely at the appropriate eye such as by employing helmets or shutter glasses.
Autostereoscopic displays do not require special viewing equipment. Such displays are attractive because they offer the closest approximation to how people view the “real” world around us, unencumbered by external viewing apparatus. Lenticular surfaces, used to simulate a 3-D image on a 2-D surface, are an example of autostereoscopic displays. The image presented on the lenticular surface typically includes at least one object and a background. Each object is associated with a relative spacing from the background, and in some cases, the other objects, so as to provide a view of the image as would appear to a moving observer of a 3-D scene. However, the present techniques for providing 3-D lenticular images are limited in that they only provide for convenient simulation of depth between objects and the background, and between the objects themselves. Moreover, the present techniques are not easily automated by, for example, a computer program. Furthermore, present techniques may only provide an adequate simulation of depth when the angle of view to the lenticular surface is changed. Accordingly, there is a need for a method for providing an image on a lenticular surface which simulates both intra- object volume and inter-object volume. There is a further need for a method of inter-object volume simulation which is adapted for automation.
Therefore, in accordance with the present invention, a series of frames are provided as part of a lenticular image. Each frame provides an animation state with certain relative displacement between objects of the image. Furthermore, each object of the image is subdivided into a plurality of layers in accordance with a mask series. The layers are displaced relative to one another in accordance relative orientation for various observation points, each corresponding to a frame in a frame sequence. Accordingly, the series of frames provides simulation of object volume by relative displacement of layers within the object. The series of frames also provides an animation sequence that includes various angle views of the object.
In one embodiment the invention provides a method for producing an input image for application from a two dimensional image onto an image surface adapted to present interlaced frames of the input image. The method receives an input image, which includes at least one object. The method receives layer definitions for a plurality of layers from at least one object of the input image. The method receives an indication of the total displacement for the object animation. The method then determines a relative displacement for each layer from the plurality of layers. The method generates a sequence of image frames by progressively offsetting said plurality of layers in accordance with relative layer displacement and a total animation displacement. Finally, the method interlaces the animated sequence of frames to a combined image to provide an input interlaced image for application to said image surface.
A number of method are employed in the prior art to create the N sub-images when 3-D information is available. For example, if a computer based 3-D model and rendering package are available, the N sub-images may be created by positioning of the simulated camera at various angles by lateral movement around the object. The same is with a physical static object that is available, where positioning of the camera is the only requirement. Using a computer model, a focal point is usually chosen in the middle of the object of interest. The computer camera position is then moved in regular steps to vary the viewing angle to the focal point. As may be appreciated, the exact step size and the number of images depends on the lens spacing of the lenticular sheets being used, the resolution of the printing system available, and the quality of the stereopis.
Often times, it is not possible to move a camera position about a subject, either by computer program or by physical relocation. For example, the object may be from a photograph taken some time before the simulation effort and is no longer available for acquiring multiple view images. At other times, the photographed subject is in motion or is continuously changing such as a splashing drink, where capturing multiple angles of the object is expensive and difficult. Furthermore, a scene may require extensive manipulation and the addition of numerous objects or may include a hand drawing for which depth information is not available. Accordingly, the present invention provides a method for simulating 3-D views of an object without acquiring images of the object at multiple angles or employing computer based 3-D model and 3-D rendering package. Rather, the invention employs a mask series and motion definitions specific to the objects to provide animated frames of the object. These animated frames are then incorporated into an interlaced image for the lenticular surface or are provided to a motion sequence viewer as an animation sequence.
The following discussion, with reference to
Each object is preferably divided to a plurality of layers, which are defined by the mask series. The mask series is preferably generated by the designer outlining and cutting away select portions of the object to define a plurality of layers. Each layer is selected so as to include portions of the object that are relatively in the same cross-section plane of the object. As may be appreciated, the object portions in a layer may include non-continuous portions which share a common plane. As illustrated below with reference to
Each frame of the animation sequence, employed to provide the interlaced lenticular image, includes a snapshot of object layers at various horizontal displacements, based on the particular observation point and corresponding focus point. The animation of object layers of step 70 in
The relative displacement value for each layer is preferably determined by reference to the desired total displacement for the object. Specifically, the relative displacement for each layer depends on its relative distance from the focus point and the maximum displacement for the layer farthest from the focus point. As may be appreciated, layers that are closer to the focus point will have a smaller relative displacement value than layers further away from the focus point. In each frame of the sequence, each layer's relative displacement is preferably a multiple of the frame number in the frame sequence. For example, in one embodiment, a layer that has a displacement of 2 pixels relative to the focus point is displaced 2 pixels in frame 1, 4 pixels in frame 2, 6 pixels in frame 3 and so forth. This displacement relative to the focus point can be referred to as a “relative displacement,” since it provides a basis displacement value which is multiplied by a frame number to arrive at a total displacement for a particular layer. As may be appreciated, the object volume is defined by the relative displacement value for each layer since the layer displacement magnitude provides the illusion of volume. Accordingly, the apparent distance between layers can be increased or decreased by changing the relative displacement values to thereby stretch or shrink the object. This allows the designer great freedom in defining objects with various volumetric characteristics.
In one embodiment, the total pixel displacement for all objects is limited to 450 pixels for positive movement (left to right) and 140 pixels for negative movement (right to left). The animation displacement preferably starts from a first frame that provides the object layers at a first extreme displacement value and ends on a final frame that provides the object layers at another extreme displacement value. The first frame position preferably includes layers displaced to an extreme left position with a maximum possible displacement of −450 pixels for positive relative displacement or +140 pixels for negative displacement. Each layer's relative displacement is inverted and multiplied by half of the total number of frames in the animation to arrive at the starting frame. The animation then proceeds to a final frame having a maximum possible displacement of +450 pixels for positive relative displacement and −140 pixels for negative relative displacement. Thus, by starting the animation from an already offset maximum position as opposed to the neutral position, the animation range is doubled to include twice the number of pixels for movement in both the positive and negative directions. Hence the animation will start from the offset maximum position and end at an offset calculated by multiplying each layer relative displacement by the total number of frames in the animation. Moreover, the animation sequence is fluid with smooth transition from a first angle to a final angle of view.
The relative displacement of the furthest layer from the focus point can be expressed as (Total displacement) divided by (Number of frames). Thus, for a symmetrical 32 frame sequence, simulating 3-D viewing, the maximum relative displacement is 280/32 or −8.75 pixels and 900/32 or +28.125 pixels. Hence the layer furthest from the focus point is limited to a relative displacement value of either −8.75 or +28.125, depending on the direction of movement. Thus, for layers selected uniformly distanced from the focus point, each layer's relative displacement can be expressed as (Maximum relative displacement) divided by (Number of layers from focus point) multiplied by (Particular layer number from focus point). Thus, for the above illustrated simulation having 21 layers, the focus point assumed to be in the center layer 11, the relative displacement is −0.875 ((280/32/10)*1) and +2.8125 ((900/32/10)*1) for layers 10 and 12, −1.75 ((140/16/10)*2) and +5.625 ((450/16/10)*2) for layers 9 and 13, and so forth. As may be appreciated, fractional displacement values can be associated with layers since the actual displacement results from multiplying by a factor, which provides a result that is later rounded to the nearest whole pixel.
The displacement of layers in each frame is preferably the relative displacement for each layer multiplied by the frame number from the focus point. Preferably, the resultant pixel displacement for each layer is rounded to the nearest whole pixel with half values rounded down. The process iterates over the entire frame sequence to provide an animation including layers of the object at various horizontal offset positions. In some embodiments, optional simulation effects are selected, such as color shift, rotation, opacity, and other filters usable in graphic software to manipulate appearance. Other optional effects include vertical displacement and vertical stretching. The animation frames are then rendered in accordance with the selected parameters.
Although the present invention was discussed in terms of certain preferred embodiments, the invention is not limited to such embodiments. A person of ordinary skill in the art will appreciate that numerous variations and combinations of the features set forth above can be utilized without departing from the present invention as set forth in the claims. Thus, the scope of the invention should not be limited by the preceding description but should be ascertained by reference to claims that follow.