TW556088B - A method and apparatus for controlling mechanism for the multi-screen display system on a virtual environment - Google Patents

Abstract

The present invention relates to a method and apparatus for controlling mechanism of a multi-screen display system on a virtual environment. Through parallel rendering technology, this invention is able to enable a PC computer group implementing multi-screen display more efficiency. The system comprises: a network transmission interface; a server being in charge of processing a view database, and up loading a completed frame data to the network transmission interface; a client receiving the frame data transferred from the network transmission interface, and implementing a 3D rendering work; and a synchronous module receiving an advance signal from the server and then transferring to the client, and receiving an acknowledge signal from the client and turning back the same to the server. Therefore, while the server is processing the frame of the view, the client is preparing the former frame, and returning back the acknowledge signal to the server through the synchronous module. The server then outputs a selected frame to the client through the synchronous module, and outputs an advance signal to the client. When the client receives the advance signal, the former frame is to be displayed on a screen, and then a next frame is to be prepared. Whereby a queue of the server can be released, and the frame processing efficiency can also be enhanced.

Description

556088 Jade, description of the invention (l) — " [Field of invention] The present invention relates to a multi-screen synchronous display system of virtual reality and its method "Go" especially to a method using parallel computing technology to make use of

The number of personal computers is large, and the display efficiency of screen computers is improved. -[Background of the invention] Virtual reality is a kind of sound and light image created by the existing technology equipment. The current technology in this area is mainly combined with computer graphics, real-time control, and multimedia sound. Light and other technologies to create a real-time response virtual reality system, using existing imaging equipment, to display visual images and stereo sound effects that can fill the human visual range, so that human vision is stimulated and feels immersive Feeling; therefore, the current general 2 series uses multiple high-resolution projection equipment to generate the ring field visual effect space 'Please refer to Figure 1, a cave-type ring field visual effect space 1 is currently located on a wall with four walls. (Projection screen material) within the enclosed space 10 ′ uses two to four projectors 丨 丨 to project the virtual screens generated by the computer 丨 2 onto the screen walls 14 via the network transmission interface 13 respectively. The images of one screen are continuous and non-overlapping, so as to form a cave space of virtual visual effects to achieve realistic effects.

The above technology was developed by the Illinois State University of Chicago E V L

Electronic Visualization Laboratory (Laboratory) —It was first published in 1992. In order to provide fast simulation effects and synchronization, the function 'the hardware equipment used is a high-level workstation-class computer and special' special hardware equipment. On the imaging system, it uses four workstation computer 12 of SiSicon Graphics Inc. as the display processing.

16665.ptd Page 6 556088 V. Description of the invention (2) " " " " '" —- Every computer 12 is connected to a projector 11 and projected on the screen f by rear projection to produce an image As a result, three computers 12 used the projection to display the left, front, and right daytime planes, and the other computer 12 was responsible for displaying the daytime planes of the floor. As a system host, computer 15 is responsible for controlling and coordinating four workstation-level drawing computers 12 'the system host has reflective memory (ref lective memory)' in order to achieve data transfer and screen synchronization update. _ Reflective memory (ref 1 ect ive memory) is a kind of shared memory technology. It uses optical fiber as a network transmission interface between computers, and quickly copies any data input from outside to other computers. Bie makes mother 0 to know that they can share the same data. 'Each computer can then use its own scene data to follow the direction of its observation point, and the corresponding scene picture is VMIC's Ref lective memc): As an example, the ry product can support a memory capacity of 64Mbyte, with a maximum of 6 computer nodes, and the transmission delay of data replication between each computer can be as low as 70 ns. Therefore, the high speed of copying of reflective memory data 'can maintain the consistency of the scene and achieve the function of visual effects synchronization. However, with the current reflective memory and optical fiber hardware equipment, the price is still too expensive, and workstation computers are not affordable for ordinary consumers; in order to provide low-cost video An empty door means someone using a group of personal computers to replace expensive workstation-grade computers and special hardware equipment. '^ Because CAVE must provide the requirements of multi-screen computer display, high-resolution day and day, 3D% scene calculation and real-time calculation. The amount of calculation has exceeded a single processing

16665.ptd

Μ Page 7 556088

The monthly dagger force is 77 two parallel solutions (Parallel Rendering) 唯一 two unique solutions. The distributed parallel computing technology is used to improve system efficiency and balance system load to achieve the function of CAVE multi-screen display. First of all, first introduce the schematic diagram of the traditional 3D computer drawing process.-Refer to Figure 2 for details, which roughly includes the following processing steps:

Yatian Display traversal 20: It is the traversal of all 3D model data in the scene. These data include 31) the model's object = (primitive), the object's normal vector and local coordinate information. In this stage f ', the 3D scene database will be used to manage the change of the scene. The scene data library will interact with the user through the input device, and change the object coordinate system (ffl) of the hall and objects through the characteristics of the 3D model. 〇de 1 c 〇〇rdinate), and perform correction tests such as collision extraction and terrain fluctuations to correct the 3D model in the scene to the correct position. Geometry sub-system phase 2 1: Object transformation 210: Convert 3D model objects to a unified World Coordinate system, screen the objects 2 1 1: Check whether the objects are within the visible range, reduce unnecessary Operation. Chroma calculation 212: Calculate the lightness of the object. ® Visual range conversion 2 1 3: Transform objects in the world coordinate system into the observation space through the visual range. Cropping (cl ipping) stage 21 4: Cut only a part of the objects in the visible range into a shape that conforms to the observation space. Conversion step 2 1 5 ·· From the observation space (V i e w s p a c e) to the screen

556088 5. Description of the invention (4) There is a perspective division step in the space (screen s p ace), which is to divide the observed space coordinates by a constant w, and use the 3-D screen correspondence to convert the coordinates to the X-Y screen coordinate system. Rasterization subsystem stage 22: The work in this stage is to eliminate the results of the previous stage by hiding the surface and describe the objects in pixels. The final image buffer (frame buf fer) will retain the resulting image . Display stage 23: The work in this stage is to output the daylight to the display screen. The client and the client show that the data has been lost; the client-to-screen should be multi-screened (cl-type. The server end is connected to the digital scene data server, and the two items on the client are tracked through the screened client. For the display steps, see the synchronization section.) The display system consists of 30 main client and server-side scenes. The traditional computer drawing process is divided into server parts, and the server side is responsible for scene change. Step 20 to the cutting stage 214. As a result of the selection at this stage, all objects outside the field of view have been removed and are responsible for the next 3D edge map work, paragraph 23. Figure 3 'is a schematic diagram of the current system using a personal computer group.

-Servo Is (server) forgive me, L-θ Μ 51 d β t architecture as its organizer 1 § 4 30 and several clients 300 are responsible for managing the scene, and merge 31, # 中 ' On the computer 310 of the feeding end 31, the network interface 32 connector end 30 maintains consistency. Client 3 1 field 30: System host 3 0 0 performs the field first All 3D models and object data construction

16665.ptd Page 9 556088 _Fifth, the invention description (5) _ retains the objects in the field of view and frame information of the scene; then, the field is coexisted into a corresponding customer On the end computer 310, the next drawing work is carried out by each customer's two-sided shell material, and the screen is displayed on the computer 310 on the screen & Since these computers 310 display the screen, each The screens are different, so the processing speed of the sample will also occur. The day and time of the computer screen will be unable to synchronize with the difference ', and the picture produced by each computer must be drawn and adjusted. Different, the complexity is also H, so it needs == and the size •, resulting in the phenomenon that the time of each computer's screen is inconsistent and time is not continuous. Therefore, in order to enable the client 3 to generate a screen mismatch 31, it must be replaced by What the server side 30 passed over, the client got up, and waited until the server side 30 sent out a synchronous ordering and storage fund π tL is now notified that the information of the daytime surface is needed, and then the update action is performed. Update the consistency of scene data between pens to make the display results correct. The fourth picture of the client 3 I is a schematic diagram of the time state of the daytime surface of the multiple tombs that are familiar with virtual reality. In the period τ, the transmission work of the IV imaging system can be summarized as follows: T comput ion at the S server · Yuancheng's calculation of the first picture. T communicat ion sends the calculation result to the continuation " Ture leg 々 ^ T singal: Notifies the drawing computer to perform the drawing operation3. Eight enjoy 5 Ji Yi body. number. Ta „’ receives the response that the graphics computer has completed the first frame, so ‘the server’

16665.ptd $ 10 pages 556088 1 5. Description of the invention ... ru) flrj · ^, A rPc〇 [nplji; i〇n + 丁 cominunication + 丁 singal + ding acknowledge ° The work of the client-side drawing computer is as follows:

Tsingal: Receive the edge map notification of the i-th daytime surface on the server side. Τ __ ^: η: Receive the scene data of the i-th day. D fender *: Perform drawing operations. Ding acknow1 edge * Sending has completed the i-th day signal. Therefore, the total number of days required for the client to complete a day-to-day period is Tsingai + Trender + Tacknowledge. According to the above ’, the total time required to complete a day and night is Tc_utiQn + 2

Tsingai + 2 TTacknc3wledge + Trender, one problem can be found from the foregoing method, that is, the i-th screen on the server must wait for the i -1 day-time display and all clients respond before starting the calculation, that is to say When the client is performing drawing operations, the server is in a waiting state and does not take full advantage of the advantages of parallel operations. [Summary of the Invention] In view of the shortcomings of the conventional technology described above, the main object of the present invention is to provide a multi-screen synchronous imaging system and method for virtual reality that can effectively solve the server-side waiting problem and accelerate the speed of drawing processing. . Another object of the present invention is to provide a multi-screen synchronous development system and method for virtual reality, which can increase the day-to-day productivity of the system. In order to achieve the above and other objectives, the main feature of the present invention is to establish a virtual reality multi-screen synchronous imaging system, which includes: a network transmission interface; and a server end, which is responsible for processing the scene database. , And upload the processed daytime data to the network transmission interface; multiple customers

16665.ptd Page 11 V. Description of the invention (8) Network transmission interface 4 〇 In this implementation, ～~ ～ transmission interface, the Ethernet system is—area, use Ethernet as the network master, 100M second The basic bandwidth is used as the pass. The communication protocol of the road is based on the server terminal 41, which is provided with a system host database (such as coordinate conversion / this host is responsible for scene correction testing), and the revised book is detected. And the shape of the view space, etc., this $ ^^^ is cut to conform to the tracking steps of the conventional drawing process of Figure 2 a step a more detailed drawing steps are not repeated here; When the scene processes the picture to the cutting stage 2 and 14 steps, the picture data is uploaded to the network transfer interface 40. After the processing is completed, the client 4 2 will be buried. In this embodiment, each client 42 Each has one; Ϊ = four clients 42, the diurnal device 420 from the network transmission interface 40, which is received by the receiver. This 3D drawing is more detailed. Perform the 3D drawing conversion steps 2 1 5 to the day-to-day display stage 2 3, ... 2 After the conversion of the conventional drawing process, the screen is displayed on the computer 42 0 = no further details; when the drawing is completed, the synchronization module 43 In this embodiment, it is the previous one. (Synchronization Β〇χ), using a synchronization box channel, can provide independent signal transmission through parallel parallel port for its communication camp, its connection method is a one-to-many shape, regardless of the amount of network data In the form of the shadow four, the input is electrically connected ^, in this embodiment, it is ~ pair 4 10, and the output is electrically connected to the host computer of the server terminal 41 to receive the server 41 The client computer of the client 42 is injured; the client 42 receives the signal sent by each client = and transmits the completion signal sent to each recorded guest code 42 to the word 16665.ptd Page 13 556088 5. Description of the Invention (ίο) After drawing the i -1 daytime surface, the picture is temporarily stored in the back buffer of the display card. In step S2, the day-to-day transmission is performed, and the server end transmits the i-th day-to-day data to each client according to the object data in the view frustum of each client. In step S3, the server receives the response signal transmitted by the client, which indicates that the client has prepared the i-1 day surface, which can be displayed on the screen. In step S4, the server sends a forward signal to the synchronization module. When the synchronization module receives the signal, it is distributed to each client. In step S5, the server executes the (i + 1) th day-to-day calculation, that is, the next day-to-day processing is performed. At the same time, after each client receives the forward signal, the i-1th day-to-day display is displayed on the screen. , And you can start to draw the i-th daytime surface. Please refer to the figure. In step S1-of performing object filtering on the server side, the following steps are included. First, the host computer must first calculate the screen data SH to be transmitted. It includes application processing operations and database update operations. When the system host is ready, Si2 waits to send data to the client s13, and then proceeds to step s2. Fig. 8 is a schematic diagram of the time state of the system during the operation of this embodiment, using the principle of parallel operation, to manage the processing time of the application and the 3D scene management (the server side processes the i-th day scene calculation) It overlaps with the computer drawing time (the action of the client drawing the i -1 picture), parallelizes the work and makes the time overlap, and the time taken for each picture execution cycle is reduced to:

16665.ptd Page 15 536088

Schematic description 215 Conversion step 22 Coloring time system 23 Daytime display phase 3 Personal computer group 30 Server end 300 Computer host 31 Client 310 Computer 3 2 Network interface 4 Multi-screen simultaneous development system of virtual reality ' 4 0 Network transmission interface 41 Server side, 410 System host 42 Plural client 43 Synchronization module 16665.ptd Page 18

Claims (1)

556088 VI. Scope of patent application 1. A multi-screen synchronous imaging system for virtual reality, including: a network transmission interface; a server end, responsible for processing the scene database, and uploading the processed screen data to the Network transmission-input interface; The client is to receive the screen data from the network transmission interface and perform the 3D drawing work; and the synchronization module can receive the previous incoming signal sent by the server and transmit To the client, and receive a completion signal from the client to the server. 2. The virtual reality multi-screen synchronous display system described in item 1 of the scope of patent application, wherein the network transmission interface is Ethernet. 3. The multi-screen synchronous display system of virtual reality as described in item 1 of the scope of patent application, wherein the network transmission interface is a fast Ethernet network. 4. The multi-screen synchronous display system of virtual reality as described in item 1 of the scope of patent application, wherein the server is provided with a system host, which is responsible for the program operation of the scene processing part. 5. The multi-screen synchronous display system of virtual reality as described in item 1 of the scope of patent application, wherein the client is provided with a personal computer device to be responsible for the program operation of the 3D drawing part. 6. The multi-screen synchronous display system of virtual reality as described in item 1 of the scope of patent application, wherein the synchronization module is a synchronization box. 7. The multi-screen synchronous display system of virtual reality according to item 6 of the scope of patent application, wherein the synchronization box has one input end and multiple output ends. 8. Simultaneous multi-screen display of virtual reality as described in item 7 of the scope of patent application 16665.ptd Page 19 556088 w 6. Application scope Patent system, where the input terminal is electrically connected to the server terminal, and each output terminal is electrically connected to the corresponding client. 9. A method for multi-screen simultaneous development of virtual reality, which includes the following steps: (1) The server performs object filtering and processes the i-frame / scenario data, and the client executes i-1 Daytime drawing work; (2) The server sends the i-th screen data to the client host through the network transmission interface, and when the client drawing work is completed, it sends a response signal to the server via the synchronization device. (3) When the server receives the response signal transmitted by the synchronization module, it sends a forward signal to notify the client via the synchronization device, and simultaneously displays i-1 screen; and (4) the client receives After the previous incoming signal transmitted by the synchronization module, the i-1 day surface is displayed on the screen, and the server side processes the scene data of the i + 1 day surface. 10. The method for multi-screen synchronous development of virtual reality as described in item 9 of the scope of patent application, wherein before step (1), a step of initializing scene information database on the server side is further provided to the server The host computer of the remote computer is in the starting state. 0-1 1. The method of multi-screen synchronous development of virtual reality described in item 9 of the scope of patent application, wherein step (1) further includes the following steps: 16665.ptd Page 20 556088 6. Scope of patent application 1. The system host on the server side first executes the processing operations of the application program and the updating operation of the data; 2. Calculates the data that needs to be transmitted to the screen; and 3. After the system finishes the calculation , That is, waiting to send data to the client. 12. The multi-screen simultaneous development method of virtual reality as described in item 9 of the scope of the patent application, wherein in step (1), after the client finishes drawing the i-1 daytime surface, the screen is temporarily Stored in the back buffer of the graphics card. 16665.ptd Page 21