Computer graphics processing and selective visual display system – Computer graphics processing – Three-dimension
Reexamination Certificate
2000-02-18
2002-08-20
Nguyen, Phu K. (Department: 2671)
Computer graphics processing and selective visual display system
Computer graphics processing
Three-dimension
Reexamination Certificate
active
06437779
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to an image processing method for processing images in real-time, an image processing method for generating an image data structure suitable for processing images in real-time, and a recording medium recording the image data structure or image processing program concerned.
2. Description of the Related Art
In a home video game machine, moving pictures of objects of a game are created in real-time synchronizing with the progress of the game. For the image processing in such a game machine, it is necessary to move the positions of the objects in the game responding to the input control signals of the operator within a short period, a frame period for example, and to draw images corresponding to the movement positions in the image memory (frame memory).
In order to express movement more naturally, it is necessary that characters, e.g. living creatures, as the objects of the game can move joint parts and muscle parts along with the movement of limbs and necks, and that the joint part becomes a different shape according to the movement of other polygons. In other words, in order to express more natural movement, the part between the upper arm and shoulder of an individual or the part between the upper arm and lower arm requires such movement as raising the muscles of the upper arm and shoulder depending on the movement of the polygons.
A drawing method for giving natural movement to such joint parts specifies the vertices to be connected for models on both sides of the joint, and adds a polygon for the joint between the vertices. With such a drawing method, however, the joint part is drawn by standardized polygons, where movement corresponding to the change of angle cannot be expressed, and as a result pictures to be generated appear unnatural.
In order to draw more natural movement, it is proposed to utilize integrated shaping models called “envelopes” using three dimensional modelers in the field of image processing of movies such as CG movies (movies using computer graphics), which do not require real-time processing. In these integrated shaping models, objects (characters) are comprised of polygons and bones which influence the polygons. In other words, an integrated shaping model is comprised of polygons, which are the outer surface of the character and are drawn based on actual movement, and bones, which influence the positions of the polygons with a predetermined weight. New positions of bones which influence the positions of the polygons are computed by a position conversion matrix determined according to the operation input of the operator, and the positions of the vertices of the polygons are determined according to the degree of influence (weight) from the bones. Complicated movements can be expressed by polygons by executing rendering (drawing) on the polygons based on the determined vertices.
FIG. 1
is a drawing depicting an example of movement of the above mentioned integrated shaping model in a CG movie. The model shown in
FIG. 1
is a model comprised of an arm part
1
and a hand part
2
, and
FIG. 1A
shows the basic form of the model. When the arm is bent such that the upper arm
3
of the arm part
1
becomes more vertical and the lower arm part
4
of the arm part
1
becomes more horizontal from the basic form, the state shown in
FIG. 1B
is expected. In other words, the joint part connecting the upper arm
3
and the lower arm part
4
exhibit a natural bent shape and the inner muscles
3
A of the upper arm
3
tense up.
FIG. 2
is an explanatory drawing depicting conventional image processing using an integrated shaping model in CG movies.
FIG. 3
is a flow chart depicting the image processing procedure thereof.
FIG. 2
shows the data structure
620
of the integrated shaping model (envelope). In this example, a total of seven models, models
1
-
7
, are in the tree type hierarchical structure shown in FIG.
2
. The data structure of each model is the same as the data
624
of the model
4
and the data
627
of the model
7
.
When a model has polygons to be displayed, the model has vertices which are the composing elements of the polygons. Therefore the data structure
624
of the model
4
includes the vertex list
632
and the polygon list
635
of the polygons comprised of a combination of the vertices. The vertex list
632
includes position information and normal line data of each vertex in model local coordinates. Each model also has a model matrix
631
, which is information denoting relative position with respect to a model in the higher hierarchy. The model
4
, for example, has rotation, shift (translation) and size (scale) as the relative position information with respect to model
3
. The model matrix
631
comprised of such position information, can also be matrix data when the coordinate system of model
3
is converted to the coordinate system of model
4
.
The data structure
624
of the model
4
also has the weight list
633
including the vertices of other models which model
4
influences, and the weight values, which indicate the degree of that influence. In the case of the example in
FIG. 1
, for example, the model of the lower arm
4
influences the muscle part
3
A of the upper arm
3
with a predetermined weight value. Therefore in the weight list of the model of the lower arm part
4
, the ID of the model of the upper arm
3
, the indexes of the vertices thereof (vertex IDs) and the weight values are included.
If the model
4
of the model structure in
FIG. 2
is the lower arm part
4
in
FIG. 1
, and the model
7
is the upper arm part
3
, then the positions of vertices of the muscle part
3
A in the upper arm part
3
model change, influenced by the movement of the upper arm part
3
and the movement of the lower arm part
4
, and the image where the muscle part
3
A rises is drawn. Now the image processing procedure by a three dimensional modeler, which is generally used for CG movies, will be explained according to FIG.
3
. At first, the matrix for implementing the movements of all models is calculated and saved in the data structure of each model (Step
602
). Then referring to the weight list
633
, vertex coordinates and normal line data (vertex data) indicated by the vertex indexes (vertex IDs) of the model (model
7
) which is influenced by the target model (model
4
) are read from the vertex list
632
of the model
7
(Step
603
), the read vertex data is converted to the vertex data when the vertex data belongs to the local coordinate system of the model
4
based on the relationship in the basic form, and the converted vertex coordinates and normal line data are multiplied by the model matrix of the model
4
(Step
604
). Also the vertex coordinates and the normal line data multiplied by the model matrix are multiplied by the weight value in the weight list
633
, and the values are integrated with the values in the weight vertex list
634
in the data structure
627
of the model
7
.
The above procedures
603
,
604
and
605
are executed for the entire weight list (Step
606
) and are executed for all models (Step
607
). As a result, coordinate conversion corresponding to the movement of characters at actual vertices of all models, and the position movement according to the degree of influence from other models are executed. Then the normal line vectors of the weight vertex list of all models are normalized so that the normal vectors can be used for light source processing in rendering, which is executed later (Step
608
). And referring to the vertex positions in the weight vertex list, the drawing (rendering) of the polygons in the polygon list is executed (Step
609
).
For the above operations, the matrix operation and weight operation are executed for the vertices of all models, and rendering processing for the polygons is executed after the above operations complete. Or, one vertex is focused on and the matrix operation and weight operation of models which influence the vertex are executed, then the new vertex coo
Ando Takashi
Saito Tomoaki
Dickstein , Shapiro, Morin & Oshinsky, LLP
Nguyen Phu K.
Sega Enterprises Ltd.
LandOfFree
Image processing method image processing unit and recording... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Image processing method image processing unit and recording..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Image processing method image processing unit and recording... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2936202