CN101303773A - Method and system for generating virtual scene - Google Patents
Method and system for generating virtual scene Download PDFInfo
- Publication number
- CN101303773A CN101303773A CNA2008101145430A CN200810114543A CN101303773A CN 101303773 A CN101303773 A CN 101303773A CN A2008101145430 A CNA2008101145430 A CN A2008101145430A CN 200810114543 A CN200810114543 A CN 200810114543A CN 101303773 A CN101303773 A CN 101303773A
- Authority
- CN
- China
- Prior art keywords
- scene
- steps
- design
- designed
- virtual
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Abstract
The invention discloses a method for generating a virtual scene and a system thereof. The method comprises the following steps of: expanding basic property information about a target scene to be designed so as to form detailed information about the target scene to be designed, hiring the detailed information about the target scene to generate a conceptive model of the target scene, instantiating the conceptive model of the target scene to obtain the target scene according to the resources for scene design. The system mainly comprises a conceptive model generating module and an instantiated module. The system can automatically design and generate virtual scene, thus greatly shortening the design period and ensuring that the design result has reusability to the utmost extent.
Description
Technical field
The present invention relates to the computer virtual technology field, particularly, relate to animation (Digital Media) manufacture technology field, particularly relate to a kind of virtual scene and generate method and system.
Background technology
In animation (Digital Media) making field, make up, from the artificial interference to the system, generate automatically to scene from space autoplacement the earliest based on semanteme based on geometrical constraint, automatically passed by the course of two more than ten years of the research that generates about scene has obtained many useful achievements.Existing work is mainly carried out around the intelligent degree and the efficient that how to improve the scene generation, and more representational system is Canadian CAPS (Constraint-based Automatic Placement for Scene Composition) system.In recent years, along with the widespread use of knowledge engineering in various fields, many researchers introduce the notion of semanteme wherein, lay particular emphasis on and make up the virtual scene with rationality and authenticity, and representative system has WordsEye and CarSim.
After this, flourish along with global range animation (Digital Media) industry, how apace, automatically make up with plot in the scene that matches of content explicit or implicit expression become a target that has more challenge.
Divide from the angle that technology and function realize, present existing work can be divided into two main directions: based on the automatic space layout of constraint with based on the automatic conversion from the text to the animation of knowledge.
Scene autoplacement system based on geometrical constraint and semantic information.This type systematic is mainly studied automatic placement and the rational deployment that how to realize internal objects in a finite space territory.With the CAPS system is example, it is by people such as the Ken Xu development of Toronto university, this system adopts a kind of automatic space layout method based on constraint, promptly based on the object space geometrical constraint in the physical vlan world and based in the real world between the object semantic constraint of position relation (as putting book on the bookshelf, desk lamp is placed on desk or tea table is first-class), automatically the placement of guiding object in the space, be convenient to the user fast, convenient and reasonably complex scene is carried out layout.Experiment shows that this system can help the user to arrange a complex space that includes 300 objects in less than 10 minutes time.
The automatic creation system of text-scene.This type systematic is input with the text of natural language, is intended to by the grammatical analysis of text and semantic understanding are converted into two-dimension picture or three-dimensional scenic automatically with the content of text.There are WordsEye and CarSim etc. in representative system.
The WordsEye system has easy and effective advantage by describe schematization information with natural language.But usually the user is detailed inadequately to the description of scene in the text of input, the knowledge of many common-senses is all concealed be not illustrated, and it is far from being enough that the information that therefore only utilizing the user is provided generates scene.WordsEye is input with the text of natural language, at first sentence is identified and analyzes, and sentence is converted to the semantic primitive that some have absolute construction, provides semantic meaning representation simultaneously; Secondly, use description rule relevant semantic meaning representation to be converted to the descriptor (can be used to express the attitude, spatial relationship, color of three-dimensional body etc.) of one group of bottom; Next, using the conduction rule clears up the constraint conflict that may exist between the situation elements and adds implicit constraint.Through above-mentioned steps, the descriptor that obtains at last can and be finished final scene in order to structure and the corresponding three-dimensional body of content of text.
The CarSim system is that this system can realize the visual of motor traffic accidents process by people such as the breadboard Sylvain Dupuy of French GREYC development.The event procedure that it is described with natural language (French) is input, the animation of this event procedure of output simulation.In information processing, at first from text, extract static information, the information of environment of living in when promptly incident takes place (as the information of the tree in road and roadside); Multidate information in the further extraction incident (as two automobiles of the litigant in the car crass incident, eyewitness, collision etc.) again.This system comprises two modules: semantic module and visualization model.The function of semantic module is by the understanding of natural language being extracted in the text descriptor with temporal and spatial correlations, natural language description being converted into a kind of formal intermediate language; Visualization model then is responsible for formal semanteme is converted into three-dimensional animation.
Conclusion is got up, and mainly there is following some deficiency in present prior art:
1. must carry out complete description to the element in the scene, promptly system can not generate the situation elements that does not occur in the description;
2. what adopt in the scene building process is the construction method of " growing out of nothing ", to representational scene example reuse and shared aspect relate to few;
3. in scene makes up, the subject behavior that less consideration wherein may occur promptly merges the scene design of behavior.
Summary of the invention
The object of the present invention is to provide a kind of virtual scene to generate method and system.It can automatically design and generate virtual scene, shortens design time greatly, and makes design result have maximum reusability.
A kind of virtual scene generation method for realizing that purpose of the present invention provides comprises the following steps:
Steps A is expanded the scene base attribute information of object scene to be designed, forms the details of object scene to be designed, utilizes the conceptual model of the details generation object scene of this object scene;
Step B carries out instantiation according to the scene design resource to the conceptual model of described object scene, generates object scene.
Comprise the following steps: before the described steps A
Steps A ', according to existing virtual scene example, obtain scene properties information, make up the scene design resource.
Also comprise the following steps: after the described step B
Step C verifies described object scene and estimates according to the scene designed regular in the described scene design resource;
Step D judges whether to satisfy the object scene requirement according to evaluation result, if not, then forwards step B to and handles; Otherwise finish behind the export target scene example.
Described steps A ' comprise the following steps:
Steps A 1 ' is chosen existing representational, virtual scene example that the value of reusing is arranged;
Steps A 2 ', from the virtual scene example, extract the material and the texture that comprise scenes function, situation elements classification, scene layout, situation elements, and the scene properties information of the different levels of scene color scheme, and the virtual scene example is portrayed and described with extendible scene concept map data structure;
Steps A 3 ', the semantic content that is contained according to scene properties information carries out semantic tagger to the scene properties information of the described different levels that extracted, and sub-category adding in the scenario resources storehouse;
Steps A 4 ', will be from described virtual scene example resulting design rule, add in the scene design knowledge base sub-categoryly.
Described steps A comprises the following steps:
Steps A 1 according to object scene to be designed, is obtained scene base attribute information and is expressed generation initial concept model with the form of predicate;
Steps A 2 is expanded described scene base attribute information according to the existing rule of object scene to be designed, forms the details of object scene to be designed;
Steps A 3 is replenished, is revised and adjust the initial concept model that is obtained by steps A 1 according to the details of resulting object scene design, obtains the conceptual model of object scene.
Described steps A 1 comprises the following steps:
Steps A 11 is imported the scene base attribute information of object scene to be designed, describes with predicate;
Steps A 12 according to the scene base attribute information of object scene to be designed, is set up the initial concept model of described object scene to be designed based on existing virtual scene example.
Described step B comprises the following steps:
Step B1, based on described scenario resources storehouse the situation elements of described object scene is carried out instantiation, situation elements object scene base attribute to be designed, that have suitable function, classification, style, material and texture is satisfied in selection, and object scene to be designed is carried out instantiation;
Step B2 according to the details of resulting object scene design, by the scene design knowledge base, carries out instantiation to spatial relation between the situation elements in the conceptual model of object scene to be designed, obtains object scene.
For realizing that purpose of the present invention also provides a kind of virtual scene generation system, be used for automatically designing and generate virtual scene, comprise conceptual model generation module and instantiation module, wherein:
Described conceptual model generation module is used for the scene base attribute information of object scene to be designed is expanded, and forms the details of object scene to be designed, utilizes these details to generate the conceptual model of object scene;
Described instantiation module is used for according to scenario resources storehouse and scene design knowledge base, and the conceptual model of described object scene is carried out instantiation, obtains object scene.
Also comprise the design evaluation module, be used for the described object scene that generates is verified and being estimated, and judge whether the design result of the object scene that generates satisfies the object scene requirement according to the scene designed regular of described scene design resource.
Described virtual scene generation system also comprises:
Load module is used to import the base attribute of object scene;
Output module is used to export the objective scene.
Also comprise human-computer interaction interface, be used for that every attribute to scene example, situation elements, material and texture, color scheme scheme information adds, deletion, editing operation.
Described scenario resources storehouse comprises scene case library, situation elements storehouse and material texture storehouse and color scheme scheme base, wherein:
The scene case library comprises a large amount of representative scene examples;
The situation elements storehouse comprises the classification of the representative situation elements that extract in a large number from the scene example;
Material texture storehouse comprises the material and the texture of the representative common elements that extract in a large number from the scene example;
The color scheme scheme base comprises the scene color scheme information with the value of reusing that extracts in a large number from the scene example.
Described scene design knowledge base comprises scene owner dependency rule, action dependency rule, event correlation rule and scene design rule and knowledge.
Beneficial effect of the present invention is:
1. at the multiple information that is contained in the scene example, adopt the method for hierarchical description to obtain and store related content, follow-uply reuse and share with what guarantee these information;
2. design knowledge, and the use of relevant empirical and common sense knowledge the correctness on guaranteeing scene geometric space layout, more can satisfy scene design result rationality, validity and diversity;
3. compare with traditional scene method for designing, the scene conceptual design is placed on design focal point on the limited content that can characterize scenes function and style, thereby makes design time shorten greatly; And, make design result have maximum reusability because of the typicalness that it had.
Description of drawings
Fig. 1 is the process flow diagram of virtual scene generation method of the present invention;
Fig. 2 is the algorithm for design of virtual scene generation method of the present invention;
Fig. 3 is the mark example of situation elements among the present invention;
Fig. 4 is the mark interface of scene example among the present invention;
Fig. 5 is the synoptic diagram of rule invocation mechanism among the present invention;
Fig. 6 is the initial concept model instance of object scene among the present invention;
Fig. 7 is the conceptual model of scene among the present invention;
Fig. 8 is based on virtual scene example and the regular cartoon scene example that generates among the present invention;
Fig. 9 is the structural representation of virtual scene generation system of the present invention;
Figure 10 is the embodiment of a scene conceptual model among the present invention;
Figure 11 is the scene according to a constructed instantiation of the described scene content of Figure 10.
Embodiment
In order to make purpose of the present invention, technical scheme and advantage clearer,, a kind of virtual scene of the present invention is generated method and system be further elaborated below in conjunction with drawings and Examples.Should be appreciated that specific embodiment described herein only in order to explanation the present invention, and be not used in qualification the present invention.
Problems of the prior art, essence are that prior art and platform only provide the scene design tool, and lack scene design example and scene design knowledge and experience.Based on this, the present invention proposes a kind of virtual scene and generate method and system.
For illustrating that virtual scene of the present invention generates method and system, at first describes the defined notion of the present invention.
1) constraint:
Constraint among the present invention is divided into two kinds of geometrical constraint and semantic constraints.Geometrical constraint mainly be for the scene layout employed; Semantic constraint then comprises the content to aspects such as scene properties, situation elements attribute and scene space formations.These constraints can make scene also satisfy relevant general knowledge (zone of action as door is to put object) when satisfying the narration requirement, and specific behavior is to the requirement (take place as the action that " seat " arranged in the scene, the height of the situation elements of being sat should equate with people's sitting height so) of scene.The present invention adopts the form of predicate to represent these constraints, as shown in Table 1 and Table 2.
The table 1 scene arrangement space geometrical constraint information content and predicate are represented
| The constraint classification | Predicate |
| The location of single body | Be positioned at coordinate points (＜situation elements name 〉, (and＜real number 〉,＜real number,＜real number 〉)); Direction (＜situation elements name 〉,＜front | the back side | end face | the bottom surface | the left side | the right side 〉,＜noun of locality〉|＜real number 〉,＜real number 〉,＜real number 〉)) |
| The location of two objects | Plan range (＜situation elements name 〉,＜left side | the right side | top | following | the front | the back 〉,＜situation elements name 〉,＜left side | the right side | top | following | the front | the back 〉,＜distance 〉) |
| Position relation between three and the above object | Between the distance (＜situation elements name 〉,＜distance,＜situation elements name 〉,＜distance,＜situation elements name 〉); The mirror image symmetry (＜situation elements name 〉,＜situation elements name 〉,＜situation elements name 〉) |
The table 2 scene layout semantic constraint information content and predicate are represented
| The constraint classification | Describing method |
| The scene properties constraint | Scene properties (＜scene name 〉, scene type,＜indoor | outdoor | the bedroom | the dining room | ...) |
| The constraint of situation elements attribute | The situation elements attribute (＜situation elements name 〉, type,＜furniture/light fixture/ornament/heating equipment/...) |
| The scene layout | Form (＜situation elements name 〉,＜situation elements name 〉) support (＜situation elements name 〉,＜situation elements name 〉) follow (＜situation elements name 〉,＜situation elements name 〉) |
2) predicate
Predicate (Predicate) is meant to be used for the speech of state, character and relation between objects of description object (persons or things).For example: the predicate form of " Xiao Ming likes reading " is---like (Xiao Ming, reading), " liking " is predicate.By predicate, the user can be fast, effectively with the present invention in virtual scene design and generation system carry out alternately automatically.
The user can describe exactly by given constraint predicate and oneself wish the picture effect obtain.
For example,, can carry out semantic description, comprise the object that should occur in the picture this picture if expect a width of cloth picture " automobile moves from the east orientation west along a path of establishing ", and spatial relation between the object etc. in the picture.
3) scene conceptual model
Described scene conceptual model is the model at the scene conceptual design.With regard to the design level, the scene conceptual design can be divided into function design and the design of personalized key element.Wherein the content of function design comprises that the classification of scene layout, situation elements is chosen and appropriate color scheme etc.; The content of personalized design then is replenishing of the personalized situation elements that is complementary with the scene master message.From reusability and sharing, the content of function design is usually can be from a large amount of design examples abstract comes out and solidifies, as newly-designed reference.Based on this, the main contents of scene conceptual design promptly are based on the relevant information of existing example and combining target scene and finish the new scene that satisfies major function and meet basic semantic constraint.
4) scene concept map data structure
What Fig. 7 represented is the scene conceptual model, and promptly any one complicated scene can take this to express.See intuitively, the multiple relation in the formation situation of scene conceptual diagram light field scape and the scene between the situation elements (as support, adjacent, relation such as follow); From intension, on behalf of in the scene, each node in the scene concept map all constitute element, and each node all has the attribute of himself, as title, material, color, physical dimension etc.Relation between the node has been represented the spatial relation that constitutes in the scene between the element.Can generate a new scene by transformation (as increasing, delete situation elements, changing the relation between the situation elements, the attribute of change situation elements etc.) to the scene concept map.So from the changeability of data structure, we the present invention is referred to as extendible scene concept map data structure.
Introduce a kind of virtual scene of the present invention generation method in detail below in conjunction with above-mentioned target, shown in Figure 2 as Fig. 1, comprise the following steps:
Step S100 according to existing relevant virtual scene example, obtains scene properties information, makes up the scene design resource;
Making up according to the scene example in the process of resources bank, in order to realize reusing and sharing of resource, need from the scene example, extract the material and the texture of representational situation elements classification and common elements respectively, and the scene properties information such as scheme of color scheme, and with these scene properties information be placed in scene case library, situation elements storehouse, material and texture storehouse, and in the color scheme scheme base.By reasonable reorganization, can obtain diversified scene design result to all kinds of design resource.
The scene design resource comprises two parts: scenario resources storehouse and scene design knowledge base.
The scenario resources storehouse comprises situation elements storehouse, scene case library, material and texture storehouse and color scheme storehouse.
Wherein:
The situation elements storehouse is the set of various situation elements, is the main composition part of scene, finally all can be transformed into operation to situation elements to all operations of scene.For the ease of realizing prompt operation to situation elements, as retrieving, reuse etc., need mark situation elements, the information of mark comprises physics, the geological information of situation elements, the semantic information that situation elements self is contained, as classification, function, the situation elements that is associated and behavior etc. and and the application scenario between possible interactive information.Be the mark interface of situation elements shown in Fig. 3 (a), Fig. 3 (b) is a situation elements part marked content.
The scene case library comprises the set of relevant with scene or akin various virtual scene examples.The virtual scene example is the foundation of virtual scene design.Existing initial virtual scene example often comprises many unordered, non-essential information, need to its screen and again the tissue, and scene properties is carried out semantic description, simultaneously the mask method of the important scenes element in the scene according to situation elements marked.Fig. 4 (a) is scene mark interface, and Fig. 4 (b) is the part marked content of scene properties.
Material and texture storehouse are to constitute the material of various necessity of scene style and the set of texture.Material and texture are the essential elementss that constitutes the scene style, in order to realize the diversified and personalized of scene design, need to collect a large amount of material and texture, and it is labelled.The content of sign also needs to add the multiple semantic informations such as feature of the style of its embodiment and suitable scene, suitable main body except the pattern content of material and texture itself is described, as Fig. 3, shown in Figure 4.
The color scheme storehouse is the set of the color information in the scene.The color scheme of scene is the color information of essential element or large tracts of land element in the scene, it is a key factor of represent scenes style, and can reflect scene owner's relevant information (as hobby, personality, economic situation, age and sex etc.) from the side.The present invention adds man-hour in that existing virtual scene example is obtained with information, choose and keep scene color scheme information with the value of reusing, and it is placed the color scheme storehouse, and and described in the mode of rule, relevant reference is provided and shares in the hope of design for new scene.
The content of scene design knowledge base comprises scene owner dependency rule, action dependency rule, event correlation rule, and empirical, common-sense scene design rule.
Among the present invention, adopted the form of production rule that knowledge is expressed.In call method to knowledge (rule), then taken all factors into consideration possible incident, behavior and scene owner's characteristic and the relation of scene design considerations in the scene, guarantee the rationality and the individual character of scene design with this.Wherein, the expression structure of production rule is as follows:
＜rule 〉: :=＜rule name 〉: IF＜regular former piece〉THEN＜regular consequent〉＜interpretation of rules〉＜regular remarks 〉
With the rule relevant with the scene integrity attribute is example, regular former piece in the scene design comprise in have (referring to Fig. 5): scene owner's physical attribute and personalization attributes, the residing epoch of scene, Domain Properties, the perhaps action that is taken place in the scene or the attribute of state-event; Regular consequent in the scene design comprise in have: the style of scenes function and scene.
The function of described scene can concern by situation elements classification and situation elements and shows.
What described scene style was concrete can be annotated by scene dominant hue, situation elements color, situation elements style and situation elements material.For example, the action that is taken place in the IF scene is for having a meal, and the function of THEN scene is have dinner (the classification default value that is scene is the dining room).
The present invention sums up the dependency rule (knowledge) of scene design from a large amount of successful scene design examples, and form a cover can be for the rule invocation method (referring to Fig. 5) of operation, the user can replenish the scene design knowledge base and expand in view of the above.
Further, the construction step of structure design resource is as follows among the described step S100:
Step S110 chooses the existing virtual scene example representational, that the value of reusing is arranged of being correlated with;
Step S120, from the virtual scene example, extract the material and the texture of scenes function, situation elements classification, scene layout, situation elements, and the scene properties information of different levels such as scene color scheme, and the scene example is portrayed and described by extendible scene concept map data structure.As shown in Figure 7.
The present invention takes the manual mode that extracts, and is described scene properties information with the defined relevant predicate form of the present invention, but is not limited to this mode.
Step S130, the semantic content that is contained according to scene properties information carries out semantic tagger to the multi-level scene properties information that is extracted among the step S120, and places different resources bank (being different situation elements storehouse, scene case library, material and texture storehouse and color scheme storehouses) respectively;
The semanteme of described implication is meant the semanteme that attribute contained of scene or situation elements, and corresponding in the scene conceptual model is exactly the semanteme that each node self is contained.
Concrete semantic content can be referring to the content shown in Fig. 3 and Fig. 4, wherein, except the geometric attribute and physical attribute (" material ", " texture ", " color ") of situation elements, also has abundant semantic attribute, as the semanteme relevant, and the scene owner semanteme of being correlated with etc. with the residing historical background of scene.Scene properties information is carried out semantic tagger, and to place the benefit of different resources banks respectively be to guarantee effectively reusing of scene example and situation elements example.
The present invention takes manual mode that scene properties information is carried out semantic tagger, but is not limited to this mode.
Step S140, will be from the scene example study to design rule and design experiences, it is added in the scene design knowledge base sub-categoryly.
The relevant scene example of scene a series of and to be designed that provides for the scene deviser among the present invention is taked artificial mode, and arrangement sums up relevant design rule and design experiences and adds to taxonomically in the scene design knowledge base from the scene example.
Step S200 according to the scene base attribute information of the object scene of importing to be designed, generates the conceptual model of object scene;
Further, described step S200 comprises the following steps:
Step S210 according to the object scene to be designed of input, obtains scene base attribute information and is expressed generation initial concept model with the form of predicate;
Step S210 comprises the following steps:
Step S211 imports the scene base attribute information of object scene to be designed, describes with predicate;
Scene base attribute information comprises scenes function and scene style information, and the information relevant with the scene owner, describes with predicate.For example:
Scene type (bedroom, bedroom)
The scene epoch (bedroom, ancient times)
Scene style (bedroom, magnificent)
Scene dominant hue (bedroom, red)
Scene owner's name (bedroom, emperor),
Owner's sex (bedroom, man)
Owner's age (growing up in the bedroom)
Owner's economic situation (bedroom, affluence)
Owner's social status (bedroom, height)
Step S212 according to the scene base attribute information of the object scene to be designed of step S211, sets up the initial concept model of described object scene to be designed based on existing relevant virtual scene example.
By the scene base attribute of being obtained among the step S211, based on the initial concept model of existing relevant virtual scene example establishing target scene.
The content of initial concept model comprises essential element classification and the basic space constituent relations thereof such as scene base attribute information that constitutes object scene to be designed, as shown in Figure 6.
Preferably, in conjunction with scene main relevant information in actual applications: action that takes place in scene owner and the scene and state, the present invention has provided the conventional method that knowledge is called (referring to Fig. 5).In the intermediate rectangular zone is the fundamental of scene design, and in right (scene owner's attribute relevant information) rectangular area, a left side (event content relevant information) is associated main information.Can derive the value of the design considerations in the middle boxes according to the value of relevant information in a left side or the right frame.
Step S220 expands the scene base attribute information of step S211 according to the existing rule of object scene to be designed, forms the details of object scene to be designed;
Usually, the user is difficult to very accurately, at large describe its all demand in input scene design requirement.Specifically can be understood as the director in the process of carrying out the video display creation, require the art designing to build the scene that corresponding story takes place according to its understanding to the story of a play or opera.In this case, the director often provides comparatively rough description and requirement, the art designing further understands in view of the above and analyzes, and according to relevant data, the observation and the experience of life are replenished the details of scene, finishes complete, a reasonably scene design at last.
When this process is simulated and realized by computer system, be that derivation by existing rule replenishes and perfect.
The flesh and blood of existing rule is the professional knowledge of scene design and designer's empirical knowledge etc.
Step S230 to being replenished, revise and adjusted by the resulting initial concept model of step S210, obtains the conceptual model of object scene according to the details that designed by the resulting object scene of step S220.As Fig. 7 and shown in Figure 10;
Figure 10 is the embodiment of a scene conceptual model, i.e. the describing mode of the scene conceptual model in a bedroom.
Step S300 carries out instantiation according to the scene design resource to the conceptual model of object scene, obtains object scene;
Further, step S300 comprises the following steps:
Step S310, based on the scenario resources storehouse situation elements of object scene is carried out instantiation, situation elements object scene base attribute to be designed, that have suitable function, classification, style, material and texture is satisfied in selection, and the situation elements in the object scene to be designed is carried out instantiation.
Step 320, according to details by the resulting object scene design of step S220, utilize the scene design knowledge base, spatial relation between the situation elements in the conceptual model of object scene to be designed is carried out instantiation, calculate object scene by relevant knowledge and physical size.
According to the described content of step S310, to situation elements size in the object scene model that is generated and the spatial relation between the situation elements, by the scene design knowledge base, carry out instantiation, selection satisfy the scene actual size, have and scenes function and the suitable size dimension of scene owner behavior and actual positional relationship and position distance.
Figure 11 is the scene according to a constructed instantiation of the described scene content of Figure 10.
Preferably, virtual scene generation method of the present invention also comprises the steps:
Step S400 is according to the empirical of the scene in scene design resource design or common-sense is regular that object scene is carried out soundness verification and evaluation;
Rule of thumb property and common sense knowledge are verified the attribute of the situation elements classification in the object scene that is generated among the step S300, situation elements and the spatial relation between the situation elements and are assessed, as have constraint conflict, then carry out conflict resolution and handle.
The method of described conflict resolution has a variety of, the conflict resolution of rule-based (knowledge) is meant under the prerequisite that guarantees geometric constraint satisfaction, according to the residing narration environment of scene its required satisfied semantic constraint is carried out priority ordering, when there being a plurality of semantic constraints, and when conflict being arranged therebetween, preferentially satisfy the high semantic constraint of importance.
Step S500 designs the object scene that generates among the step S300 once more according to evaluation result, judges whether to satisfy the object scene requirement, if not, then forwards step S300 to and handles; Otherwise finish behind the export target scene example.As shown in Figure 8.
Corresponding to a kind of virtual scene generation method of the present invention, the present invention also provides a kind of virtual scene generation system, and as shown in Figure 9, it comprises:
Scenario resources storehouse 40 comprises scene case library, situation elements storehouse and material texture storehouse and color scheme scheme base; It comprises the space geometry information and the semantic information of abundant, relevant with scene design scene inscape; Be according to existing relevant virtual scene example, obtain scene properties information, make up generation.
Described scene case library 41 comprises the set of relevant with scene or akin various virtual scene examples.
Described situation elements storehouse 42 is set of various situation elements, is the main composition part of scene, finally all can be transformed into operation to situation elements to all operations of scene.
Described material texture storehouse 43 is to constitute the material of various necessity of scene style and the set of texture.
Described color scheme scheme base 44 is set of the color information in the scene.The color scheme of scene is the color information of essential element or large tracts of land element in the scene, it is a key factor of represent scenes style, and can reflect scene owner's relevant information (as hobby, personality, economic situation, age and sex etc.) from the side.
Scene design knowledge base 50 comprises scene owner dependency rule, action dependency rule, event correlation rule and scene design rule and empirical, common sense knowledge;
Conceptual model generation module 60 is used for the scene base attribute information according to the object scene to be designed of input, generates the conceptual model of object scene.
Preferably, described virtual scene generation system also comprises human-computer interaction interface, is used for operations such as every attribute to scene example, situation elements, material and texture, color scheme scheme information adds, deletion, editor.The object scene that it generates for system, system also provides a kind of convenient, fast, friendly human-computer interaction interface, being convenient to the user can be intuitively, easily the every attribute to scene example, situation elements, material and texture, color scheme scheme information add, operation such as deletion, editor.
Virtual scene generation system of the present invention is handled virtual scene with the process identical with virtual scene generation method, therefore, describes in detail no longer one by one in the present invention.
Beneficial effect of the present invention is:
1. at the multiple information that is contained in the scene example, adopt the method for hierarchical description to obtain and store related content, follow-uply reuse and share with what guarantee these information;
2. design knowledge, and the use of relevant empirical and common sense knowledge the correctness on guaranteeing scene geometric space layout, more can satisfy scene design result rationality, validity and diversity;
3. compare with traditional scene method for designing, the scene conceptual design is placed on design focal point on the limited content that can characterize scenes function and style, thereby makes design time shorten greatly; And, make design result have maximum reusability because of the typicalness that it had.
In conjunction with the drawings to the description of the specific embodiment of the invention, others of the present invention and feature are conspicuous to those skilled in the art.
More than specific embodiments of the invention are described and illustrate it is exemplary that these embodiment should be considered to it, and be not used in and limit the invention, the present invention should make an explanation according to appended claim.
Claims (13)
1, a kind of virtual scene generation method is used for automatically designing and generate virtual scene, it is characterized in that, comprises the following steps:
Steps A is expanded the base attribute information of object scene to be designed, forms the details of object scene to be designed, utilizes the conceptual model of the details generation object scene of this object scene;
Step B carries out instantiation according to the scene design resource to the conceptual model of described object scene, generates object scene.
2, virtual scene generation method according to claim 1 is characterized in that, comprises the following steps: before the described steps A
Steps A ', according to existing virtual scene example, obtain scene properties information, make up the scene design resource.
3, virtual scene generation method according to claim 2 is characterized in that, also comprises the following steps: after the described step B
Step C verifies described object scene and estimates according to the scene designed regular in the described scene design resource;
Step D judges whether to satisfy the object scene requirement according to evaluation result, if not, then forwards step B to and handles; Otherwise finish behind the export target scene example.
4, virtual scene generation method according to claim 2 is characterized in that, described steps A ' comprise the following steps:
Steps A 1 ' is chosen existing representational, virtual scene example that the value of reusing is arranged;
Steps A 2 ', from the virtual scene example, extract the material and the texture that comprise scenes function, situation elements classification, scene layout, situation elements, and the scene properties information of the different levels of scene color scheme, and the virtual scene example is portrayed and described with extendible scene concept map data structure;
Steps A 3 ', the semantic content that is contained according to scene properties information carries out semantic tagger to the scene properties information of the described different levels that extracted, and sub-category adding in the scenario resources storehouse;
Steps A 4 ', will be from described virtual scene example resulting design rule, add in the scene design knowledge base sub-categoryly.
5, according to each described virtual scene generation method of claim 1 to 3, it is characterized in that described steps A comprises the following steps:
Steps A 1 according to object scene to be designed, is obtained scene base attribute information and is expressed generation initial concept model with the form of predicate;
Steps A 2 is expanded described scene base attribute information according to the existing rule of object scene to be designed, forms the details of object scene to be designed;
Steps A 3 is replenished, is revised and adjust the initial concept model that is obtained by steps A 1 according to the details of resulting object scene design, obtains the conceptual model of object scene.
6, virtual scene generation method according to claim 5 is characterized in that, described steps A 1 comprises the following steps:
Steps A 11 is imported the scene base attribute information of object scene to be designed, describes with predicate;
Steps A 12 according to the scene base attribute information of object scene to be designed, based on existing virtual scene example, is set up the initial concept model of described object scene to be designed.
7, virtual scene generation method according to claim 5 is characterized in that, described step B comprises the following steps:
Step B1, based on described scenario resources storehouse the situation elements of described object scene is carried out instantiation, situation elements object scene base attribute to be designed, that have suitable function, classification, style, material and texture is satisfied in selection, and object scene to be designed is carried out instantiation;
Step B2 according to the details of resulting object scene design, by the scene design knowledge base, carries out instantiation to spatial relation between the situation elements in the conceptual model of object scene to be designed, obtains object scene.
8, a kind of virtual scene generation system is used for automatically designing and generate virtual scene, it is characterized in that, comprises conceptual model generation module and instantiation module, wherein:
Described conceptual model generation module is used for the scene base attribute information of object scene to be designed is expanded, and forms the details of object scene to be designed, utilizes the conceptual model of the details generation object scene of this object scene;
Described instantiation module is used for according to scenario resources storehouse and scene design knowledge base, and the conceptual model of described object scene is carried out instantiation, obtains object scene.
9, virtual scene generation system according to claim 8, it is characterized in that, also comprise the design evaluation module, be used for the object scene that generates is verified and being estimated, and judge whether the design result of the object scene that generates satisfies the object scene requirement according to the scene designed regular of described scene design resource.
10, according to Claim 8 or 9 described virtual scene generation systems, it is characterized in that, also comprise:
Load module is used to import the base attribute of object scene;
Output module is used to export the objective scene.
11, virtual scene generation system according to claim 10, it is characterized in that, also comprise human-computer interaction interface, be used for that every attribute to scene example, situation elements, material and texture, color scheme scheme information adds, deletion, editing operation.
12, according to Claim 8 or 9 described virtual scene generation systems, it is characterized in that described scenario resources storehouse comprises scene case library, situation elements storehouse and material texture storehouse and color scheme scheme base, wherein:
The scene case library comprises a large amount of representative scene examples;
The situation elements storehouse comprises the classification of the representative situation elements that extract in a large number from the scene example;
Material texture storehouse comprises the material and the texture of the representative common elements that extract in a large number from the scene example;
The color scheme scheme base comprises the scene color scheme information with the value of reusing that extracts in a large number from the scene example.
13, virtual scene generation system according to claim 8 is characterized in that, described scene design knowledge base comprises scene owner dependency rule, action dependency rule, event correlation rule and scene design rule and knowledge.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA2008101145430A CN101303773A (en) | 2008-06-10 | 2008-06-10 | Method and system for generating virtual scene |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA2008101145430A CN101303773A (en) | 2008-06-10 | 2008-06-10 | Method and system for generating virtual scene |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101303773A true CN101303773A (en) | 2008-11-12 |
Family
ID=40113665
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNA2008101145430A Pending CN101303773A (en) | 2008-06-10 | 2008-06-10 | Method and system for generating virtual scene |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101303773A (en) |
Cited By (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102004963A (en) * | 2010-12-02 | 2011-04-06 | 张磊 | Digital city generation method and system |
| CN102073498A (en) * | 2010-12-28 | 2011-05-25 | 北京像素软件科技股份有限公司 | Method for constructing virtual scene |
| CN102231210A (en) * | 2011-07-15 | 2011-11-02 | 长安大学 | Accident analysis and simulated reconstruction computer system of cars out of control bumping onto fixtures |
| CN102541448A (en) * | 2010-12-21 | 2012-07-04 | 微软公司 | Human interactive proofs leveraging virtual techniques |
| CN101877144B (en) * | 2009-11-27 | 2012-07-18 | 深圳职业技术学院 | Three-dimensional modeling method of vast books of virtual library |
| CN102609985A (en) * | 2012-02-29 | 2012-07-25 | 北京恒泰实达科技发展有限公司 | Three-dimensional power station graphic platform |
| CN102663799A (en) * | 2010-12-24 | 2012-09-12 | 达索系统公司 | Creation of a playable scene with an authoring system |
| CN102819858A (en) * | 2012-07-30 | 2012-12-12 | 北京中科盘古科技发展有限公司 | Method for organizing and applying cartoon material |
| CN103105929A (en) * | 2013-01-05 | 2013-05-15 | 北京农业信息技术研究中心 | Interactive design and experience method of virtual agricultural park and system |
| CN103136781A (en) * | 2011-11-30 | 2013-06-05 | 国际商业机器公司 | Method and system of generating three-dimensional virtual scene |
| CN105474268A (en) * | 2014-03-27 | 2016-04-06 | 吉欧技术研究所股份有限公司 | Image display system |
| CN106709985A (en) * | 2016-12-21 | 2017-05-24 | 民政部零研究所 | Generation method and device for virtual three-dimensional sacrifice scene |
| CN109409255A (en) * | 2018-10-10 | 2019-03-01 | 长沙千博信息技术有限公司 | A kind of sign language scene generating method and device |
| CN109584354A (en) * | 2018-11-06 | 2019-04-05 | 重庆邮电大学 | It is a kind of text scape conversion in interactive scene generation method |
| CN110427227A (en) * | 2019-06-28 | 2019-11-08 | 广东虚拟现实科技有限公司 | Generation method, device, electronic equipment and the storage medium of virtual scene |
| CN111158480A (en) * | 2019-12-27 | 2020-05-15 | 于斌 | Scene model sharing method and system, augmented reality device and storage medium |
| CN112396677A (en) * | 2020-11-25 | 2021-02-23 | 武汉艺画开天文化传播有限公司 | Animation production method, electronic device, and storage medium |
| CN112784002A (en) * | 2021-02-07 | 2021-05-11 | 腾讯科技（深圳）有限公司 | Virtual scene generation method, device, equipment and storage medium |
| CN113256802A (en) * | 2021-06-17 | 2021-08-13 | 中山大学 | Virtual three-dimensional reconstruction and scene creation method for building |
| CN114154038A (en) * | 2021-11-02 | 2022-03-08 | 绘见科技(深圳)有限公司 | Method and device for pushing virtual content information in batches, computer equipment and storage medium |
-
2008
- 2008-06-10 CN CNA2008101145430A patent/CN101303773A/en active Pending
Cited By (31)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101877144B (en) * | 2009-11-27 | 2012-07-18 | 深圳职业技术学院 | Three-dimensional modeling method of vast books of virtual library |
| CN102004963B (en) * | 2010-12-02 | 2013-04-10 | 张磊 | Digital city generation method and system |
| CN102004963A (en) * | 2010-12-02 | 2011-04-06 | 张磊 | Digital city generation method and system |
| CN102541448A (en) * | 2010-12-21 | 2012-07-04 | 微软公司 | Human interactive proofs leveraging virtual techniques |
| CN102663799A (en) * | 2010-12-24 | 2012-09-12 | 达索系统公司 | Creation of a playable scene with an authoring system |
| CN102663799B (en) * | 2010-12-24 | 2017-05-17 | 达索系统公司 | Method and device for creation of a playable scene with an authoring system |
| CN102073498A (en) * | 2010-12-28 | 2011-05-25 | 北京像素软件科技股份有限公司 | Method for constructing virtual scene |
| CN102231210A (en) * | 2011-07-15 | 2011-11-02 | 长安大学 | Accident analysis and simulated reconstruction computer system of cars out of control bumping onto fixtures |
| CN102231210B (en) * | 2011-07-15 | 2013-03-06 | 长安大学 | Accident analysis and simulated reconstruction computer system of cars out of control bumping onto fixtures |
| CN103136781B (en) * | 2011-11-30 | 2016-06-08 | 国际商业机器公司 | For generating method and the system of three-dimensional virtual scene |
| US11574438B2 (en) | 2011-11-30 | 2023-02-07 | International Business Machines Corporation | Generating three-dimensional virtual scene |
| CN103136781A (en) * | 2011-11-30 | 2013-06-05 | 国际商业机器公司 | Method and system of generating three-dimensional virtual scene |
| US11069130B2 (en) | 2011-11-30 | 2021-07-20 | International Business Machines Corporation | Generating three-dimensional virtual scene |
| CN102609985A (en) * | 2012-02-29 | 2012-07-25 | 北京恒泰实达科技发展有限公司 | Three-dimensional power station graphic platform |
| CN102819858A (en) * | 2012-07-30 | 2012-12-12 | 北京中科盘古科技发展有限公司 | Method for organizing and applying cartoon material |
| CN102819858B (en) * | 2012-07-30 | 2015-07-01 | 北京中科盘古科技发展有限公司 | Method for organizing and applying cartoon material |
| CN103105929A (en) * | 2013-01-05 | 2013-05-15 | 北京农业信息技术研究中心 | Interactive design and experience method of virtual agricultural park and system |
| CN105474268A (en) * | 2014-03-27 | 2016-04-06 | 吉欧技术研究所股份有限公司 | Image display system |
| CN106709985A (en) * | 2016-12-21 | 2017-05-24 | 民政部零研究所 | Generation method and device for virtual three-dimensional sacrifice scene |
| CN106709985B (en) * | 2016-12-21 | 2020-07-14 | 民政部一零一研究所 | Virtual sacrifice three-dimensional scene generation method and device |
| CN109409255A (en) * | 2018-10-10 | 2019-03-01 | 长沙千博信息技术有限公司 | A kind of sign language scene generating method and device |
| CN109584354A (en) * | 2018-11-06 | 2019-04-05 | 重庆邮电大学 | It is a kind of text scape conversion in interactive scene generation method |
| CN110427227B (en) * | 2019-06-28 | 2023-01-06 | 广东虚拟现实科技有限公司 | Virtual scene generation method and device, electronic equipment and storage medium |
| CN110427227A (en) * | 2019-06-28 | 2019-11-08 | 广东虚拟现实科技有限公司 | Generation method, device, electronic equipment and the storage medium of virtual scene |
| CN111158480A (en) * | 2019-12-27 | 2020-05-15 | 于斌 | Scene model sharing method and system, augmented reality device and storage medium |
| CN112396677A (en) * | 2020-11-25 | 2021-02-23 | 武汉艺画开天文化传播有限公司 | Animation production method, electronic device, and storage medium |
| CN112784002A (en) * | 2021-02-07 | 2021-05-11 | 腾讯科技（深圳）有限公司 | Virtual scene generation method, device, equipment and storage medium |
| CN112784002B (en) * | 2021-02-07 | 2023-09-26 | 腾讯科技（深圳）有限公司 | Virtual scene generation method, device, equipment and storage medium |
| CN113256802A (en) * | 2021-06-17 | 2021-08-13 | 中山大学 | Virtual three-dimensional reconstruction and scene creation method for building |
| CN114154038A (en) * | 2021-11-02 | 2022-03-08 | 绘见科技(深圳)有限公司 | Method and device for pushing virtual content information in batches, computer equipment and storage medium |
| CN114154038B (en) * | 2021-11-02 | 2024-01-19 | 绘见科技(深圳)有限公司 | Virtual content information batch pushing method and device, computer equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101303773A (en) | Method and system for generating virtual scene | |
| US7707486B2 (en) | Logical spreadsheets | |
| Tutenel et al. | Generating consistent buildings: a semantic approach for integrating procedural techniques | |
| CN102193782B (en) | Method for operating table model | |
| Thalheim et al. | Sitelang: Conceptual modeling of internet sites | |
| CN106447786A (en) | Parallel space establishing and sharing system based on virtual reality technologies | |
| CN108228984A (en) | Fit up project management system, its method, electronic equipment and computer program product | |
| Prakash et al. | Rethinking global modernism: Architectural historiography and the postcolonial | |
| Amoruso | Putting Tradition Into Practice: Heritage, Place and Design: Proceedings of 5th INTBAU International Annual Event | |
| Zhang et al. | ICA3D–Intelligent computer-aided ancient Chinese architecture design | |
| Özkan | Cool Istanbul: Urban enclosures and resistances | |
| Costagliola et al. | An approach for authoring 3D cultural heritage exhibitions on the web | |
| Hardy | Map production from an active object database, using dynamic representation and automated generalisation | |
| Giordano et al. | Representation Challenges: Searching for New Frontiers of AR and AI Research | |
| Kontakis et al. | Applying aesthetic rules in virtual environments by means of Semantic Web technologies | |
| Jang et al. | Content curation for spatial experience of architectural heritage | |
| Reitz et al. | A processing pipeline for X3D earth-based spatial data view services | |
| Li et al. | ContextMap: modeling scenes of the real world for context-aware computing | |
| Nelson | What Is a Landscape? | |
| Zurndorfer | OLD AND NEW VISIONS OF MING SOCIETY AND CULTURE. | |
| Poelman | The Cob-A Parametric approach for designing affordable & high density housing: A study towards improving affordable and high density housing, using the advantages of computerization in architectural practice | |
| Contin et al. | Unveiling Beauty Through Maps Affective Image Determination for Spatial Learning Through Metropolitan Cartography Maps. | |
| Romberg | Gan’s Constructivism | |
| Barret et al. | Lessons in user interface design in the procedural city generation for games tool Ürban PAD | |
| Aksamija | Computational representations of architectural design for tall buildings |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C12 | Rejection of a patent application after its publication | ||
| RJ01 | Rejection of invention patent application after publication |
Open date: 20081112 |