Computer graphics processing and selective visual display system – Computer graphics processing – Three-dimension
Reexamination Certificate
1998-09-08
2001-07-03
Vo, Cliff N. (Department: 2772)
Computer graphics processing and selective visual display system
Computer graphics processing
Three-dimension
Reexamination Certificate
active
06256036
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates generally to three-dimensional modeling apparatuses and methods therefor, media storing three-dimensional modeling programs, and media storing three-dimensional estimation programs, and more particularly, to a three-dimensional modeling apparatus for producing a three-dimensional model of an object of interest, a method therefor, and a medium storing three-dimensional modeling programs.
2. Description of the Background Art
In general, in order to produce a three-dimensional graphics image, a three-dimensional model expressing the shape and texture (color and patterns) of a three-dimensional object of interest is produced.
Conventional three-dimensional modeling methods for producing three-dimensional models include for example silhouette and stereo methods. In the silhouette method, an object of interest is shot by monocular cameras from a plurality of visual points, and a three-dimensional model of the object is produced based on a plurality of resulting silhouettes.
In the silhouette method, however, a three-dimensional model is produced based on the contour of the shot object, any recessed part of the object is hardly expressed.
A method of estimating the shape of an object of interest by voting color information is disclosed by “Estimation of Object Shape and Color by Bidirectional Voting Using Continuous Color Images”, Hisayoshi Zaiman et al., 53rd National Meetings of Information Processing Society of Japan (The 2nd half of 1996), pp 2-301 and 302. The method is based on the assumption that the same color should be voted in a surface Voxel. According to the method, however, color voting is conducted using all the image information, and image information in the backside direction is also voted for a surface Voxel. As a result, different colors are usually voted for a surface Voxel. In order to solve this problem, color candidates voted for a Voxel should be selected. For the selection process, the method requires a large amount of memories for retaining all the colors voted. In addition, in order to determine a surface Voxel by extracting the identity of colors among the candidates, the same part should be always shot by the same color. Alternatively, a threshold should be set for determining the sameness of colors. Thus setting the threshold generally involves difficult processes.
SUMMARY OF THE INVENTION
It is an object of the present invention to provide a simple and inexpensive three-dimensional modeling apparatus capable of expressing a recessed part of an object of interest and a method therefor.
According to the present invention, a three-dimensional modeling apparatus for producing a three-dimensional model of an object of interest includes an object image capturing device, a shape estimation unit, and a producing unit. The object image capturing device shoots an object of interest from a plurality of visual points to obtain a plurality of object images. The shape estimation unit estimates the shape of the object of interest based on dispersion in image information between at least two object images among the plurality of object images. The producing unit produces a three-dimensional model based on the estimated shape.
Preferably, the image information of the object images is on color and/or density. Here, density is the intensity of a pixel of a gray-scale image.
The shape estimation unit preferably includes an image information accumulation unit and a calculation unit. The image information accumulation unit accumulates corresponding image information on at least two object images in each region of a three-dimensional virtual space formed by a plurality of regions, more specifically, in a storage region corresponding to each region. The calculation unit calculates dispersion in the image information accumulated in the region.
The three-dimensional modeling apparatus preferably further includes a presence region estimation unit. The presence region estimation unit approximately estimates a region in which an object of interest is present in the three-dimensional virtual space based on the contour of the object in the plurality of object images. The image information accumulation unit accumulates image information in a region within the presence region among the region of the three-dimensional virtual space. Herein, the image information is preferably accumulated only in a region within the presence region.
The shape estimation unit preferably further includes a coupling unit. The coupling unit smoothly couples regions with small dispersion among the regions of the three-dimensional virtual space, in order to estimate the shape of the object.
The coupling unit preferably minimizes an evaluation function using as a variable the smallness of dispersion and the smoothness of regions to be coupled, in order to smoothly couple regions with small dispersion.
The shape estimation unit preferably further includes a coordinate transformation unit and a coupling unit. The coordinate transformation unit transforms one coordinate space visual point in a three-dimensional virtual space into a perspective coordinate space having radial coordinate axes. The coupling unit couples regions having the minimum dispersion or dispersion larger than the minimum value by a prescribed value among regions on a straight line passing through one visual point in the perspective coordinate space. The coordinate space of the three-dimensional virtual space is a rectangular coordinate space.
The shape estimation unit preferably further includes a coupling unit. The coupling unit couples regions having the minimum dispersion or dispersion larger than the minimum value by a prescribed value among regions on a straight line passing through the visual points in a three-dimensional virtual space, in order to estimate the shape of an object of interest. The coordinate space in the three-dimensional virtual space is further preferably a rectangular coordinate space.
The shape estimation unit preferably includes a partial surface estimation unit and an integration unit. The partial surface estimation unit estimates a plurality of partial surfaces (virtual shapes) of an object of interest based on dispersion in corresponding image information between a plurality of object images. The integration unit integrates the plurality of partial surfaces to determine the shape of the object.
The three-dimensional modeling apparatus preferably further includes a presence region estimation unit. The presence region estimation unit approximately estimates a region in which an object of interest is present in the three-dimensional virtual space based on the contour of the object in the plurality of object images. The shape estimation unit preferably includes an image information accumulation unit, a calculation unit, a partial surface estimation unit and an integration unit. The image information accumulation unit accumulates corresponding image information of at least two object images in each region of a three-dimensional virtual space formed by a plurality of regions, more specifically in a storage region corresponding to each region. The calculation unit calculates dispersion in the image information accumulated in the region. The partial surface estimation unit estimates a plurality of partial surfaces (virtual shapes) of the object of interest based on dispersion in the corresponding image information between the plurality of object images. The integration unit integrates the plurality of partial surfaces to determine the shape of the object. The integration unit includes a distance accumulation unit and a presumption unit. The distance accumulation unit accumulates distances to the plurality of partial surfaces from a region inside the presence region and outside each of the partial surfaces among the regions of the three-dimensional virtual space in the region inside the presence region and outside each of the partial surfaces, more specifically from a storage region corresponding to each region, in that storage region. The pr
Armstrong Westerman Hattori McLeland & Naughton LLP
Sanyo Electric Co,. Ltd.
Vo Cliff N.
LandOfFree
Three-dimensional modeling apparatus for producing... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Three-dimensional modeling apparatus for producing..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Three-dimensional modeling apparatus for producing... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2564492