CN102591709B - Shapefile master-slave type parallel writing method based on OGR (open geospatial rule) - Google Patents
Shapefile master-slave type parallel writing method based on OGR (open geospatial rule) Download PDFInfo
- Publication number
- CN102591709B CN102591709B CN201110441737.3A CN201110441737A CN102591709B CN 102591709 B CN102591709 B CN 102591709B CN 201110441737 A CN201110441737 A CN 201110441737A CN 102591709 B CN102591709 B CN 102591709B
- Authority
- CN
- China
- Prior art keywords
- shapefile
- data
- file
- layer
- processing
- 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.)
- Expired - Fee Related
Links
- 238000000034 method Methods 0.000 title claims abstract description 163
- 230000008569 process Effects 0.000 claims abstract description 124
- 238000012545 processing Methods 0.000 claims abstract description 46
- 238000001983 electron spin resonance imaging Methods 0.000 claims description 5
- 230000005540 biological transmission Effects 0.000 claims description 4
- 125000004122 cyclic group Chemical group 0.000 claims description 2
- 238000004364 calculation method Methods 0.000 abstract description 2
- 238000000638 solvent extraction Methods 0.000 abstract 1
- 230000006870 function Effects 0.000 description 16
- 238000004891 communication Methods 0.000 description 5
- 238000011161 development Methods 0.000 description 5
- 238000011160 research Methods 0.000 description 5
- 238000012360 testing method Methods 0.000 description 5
- 238000004422 calculation algorithm Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- 238000003672 processing method Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 239000003054 catalyst Substances 0.000 description 2
- 239000000203 mixture Substances 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- 244000025254 Cannabis sativa Species 0.000 description 1
- 230000001133 acceleration Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000013475 authorization Methods 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 238000004883 computer application Methods 0.000 description 1
- 238000013501 data transformation Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 230000013011 mating Effects 0.000 description 1
- 238000007639 printing Methods 0.000 description 1
- 238000004886 process control Methods 0.000 description 1
- 238000011112 process operation Methods 0.000 description 1
Images
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Multi Processors (AREA)
Abstract
The invention belongs to the field of high-performance geographic calculation, and discloses a shapefile master-slave type parallel writing method based on the OGR (open geospatial rule), which includes: step 1, inputting command line parameters; step 2, setting up a shapefile target data source and a layer file serially, and closing a target data source file; step 3, opening a shapefile original file to be processed to acquire each layer and graphic data volumes of the shapefile; step 4, performing MPI (message passing interface) parallel initialization to acquire numbers and quantity of processes, and setting a master process and slave processes; step 5, performing data partitioning to determine initial fid and terminal fid of graphics in the shafpefile processed by each process; and step 6, allowing each slave process to perform shapefile data processing. Computing resources are fully used, and integral efficiency of shapefile processing is improved.
Description
Technical field
The invention belongs to high-performance geocomputation field, particularly related to a kind of shapefile file Slave Parallel write method based on OGR.
Background technology
Rapid progress along with earth observation technology, magnanimity multidimensional space-time data is growing, geography information becomes increasingly abundant, geographical vector data writes the feature that disk more and more presents highly dense after processing, and shapefile (a kind of form of the geometry of the description spatial data of ESRI company and the non-topological entity vector data structure of attributive character) file is the vector data file the most often using.The serial WriteMode of existing Geographic Information System and conventional hardware platform, be difficult to support the demand of the fast literary sketch of magnanimity shapefile file.Universal gradually along with the New Hardware framework of parallel computing trunking and polycaryon processor, provides opportunity for being limited by the fast literary sketch of shapefile file that calculated performance is difficult to carry out.Therefore, invent efficient shapefile file in parallel write method have important researching value based on New Hardware framework, it will lay the foundation for magnanimity shapefile data processing, large-scale complex geocomputation.
OGR is GDAL(Geospatial Data Abstraction Library) the Yi Ge branch of project, function and GDAL are similar, and only it provides the support to vector data.GDAL is that a name in X/MIT(MIT permission agreement is derived from (the Massachusetts Institute of Technology of Massachusetts Institute of Technology (MIT), MIT), claiming again " X permission agreement " (X License), is in many soft ware authorization clauses, and what be widely used is wherein a kind of.) the raster spatial data transformation warehouse of increasing income under permission agreement.Having a lot of famous GIS(Geographic Information System) series products all used GDAL/OGR storehouse, the ArcGIS9.2 that comprises ESRI, Google Earth and cross-platform GRASS GIS (Geographic Resources Analysis Support System) system etc., the present invention has equally also used the built-in function of OGR.
In the development history of computing machine, the approach that improves computing machine processing speed has two, and the one, improve components and parts to improve its performance, the 2nd, the concurrency of development system.In recent years, due to PC(Personal Pomputer) appearance of the increasing substantially of machine arithmetic speed, the reduction of hardware price and the PC of unit multi-core CPU, the processing speed of computing machine has not been problem, and how to utilize efficiently these high performance computing machines to become the focus of research, MPI(Message Passing Interface, one of standard of message-passing parallel program design) development of parallel computation programming development software engineering, for parallel computation problem provides new thinking.MPI is that a parallel computation message passing interface ，You MPI forum releases, and the object of formulating this standard is to improve portability and the development efficiency of concurrent program.2009, Lu Yutong, Yang Xuejun etc. have delivered " a kind of parallel computation method for allocating tasks towards multiple nucleus system " literary composition in periodical < < computer research and progress > > the 46th volume Suppl. version, set up multi-core processor oriented progress scheduling model, designed and Implemented parallel task to multinuclear mapping algorithm, solved better the Task Allocation Problem of multi-core processor oriented in extensive resource management system, phase mutual interference while having reduced a plurality of process operation, promoted application program capacity.2004, Lv Jie etc. are infrared and laser engineering > > 33 volumes the 5th interim delivering " MPI parallel computation is in the application of an image processing method face " literary composition at < <, discussed emphatically the Parallel Implementation method of image processing algorithm, enumerated some and improved some measures of parallel algorithm efficiency.2010, Zeng Yan delivers " research and implementation that the master-slave mode task based on MPI is distributed " at < < Computer application for the 6th phase of software > >, by carrying out task with connected component algorithm, distribute to realize load balance, thereby minimizing program execution time, has improved counting yield.Master-slave mode is a kind of more complicated pattern of MPI parallel Programming, need to carry out to principal and subordinate's process division and principal and subordinate's interprocess communication Strategy Design of function.Forefathers' research stresses the parallelization research of data processing method more, and do not consider when data volume is very large, the I/O of data and disk (input/output) time will have a strong impact on the efficiency of program execution, and wherein in vector data a large amount of geodata mainly with the storage of shapefile file, so significant to the efficient processing of shapefile file to the invention of the concurrent write method of shapefile file.
The invention of Shapefile file in parallel write method will significantly improve the demand of magnanimity geographical space shapefile file high-performance under the efficiency of every profession and trade geographic information application, particularly regional scale, enterprise-level application, high-level efficiency, high-quality processing.The invention of the geographical shapefile file of magnanimity high performance parallel write method, meets current techniques development trend completely, has researching value and practical value.
Summary of the invention
1. the technical matters that invention will solve
Method of the present invention is difficult to support the deficiency of the demand of the fast literary sketch of magnanimity shapefile file for the serial WriteMode of existing Geographic Information System and conventional hardware platform, a kind of shapefile file Slave Parallel write method based on OGR is provided.
2. technical scheme of the present invention is as follows:
Principle: Slave Parallel method is to control each by host process to complete and distribute to each from the task of process from process.OGR is model target data source to the process of writing of shapefile file, then sets up the figure layer of target data source, and the graph data in the most former shapefile file writes figure layer.The present invention is that first serial creates target data source and figure layer file, then adopts Slave Parallel method, by host process, controls from process and according to arriving first the strategy of first writing, to shapefile target data source, writes graph data successively.The data volume of the shapefile file that each is processed from process is divided according to No. FID of the quantity from process and former data (numbering to figure in figure floor shapefile file).Each is to write in the mode to existing file to add data from process to writing of target data source file.
Shapefile file Slave Parallel write method based on OGR, comprises the following steps (concrete steps process flow diagram is shown in Fig. 1):
Step 1: utility command line program fill order (mpirun), and select the process number needing to carry out this concurrent write program.Input command line parameter: mpirun-np8hpgc_para_O-l Changsha～data changsha.shp～data test_result.img.
Wherein, " np " " l " is the boot symbol of order line, and " mpirun " represents to call the statement parameter of MPI application program; " n p8 " represents to call 8 process numbers; " hpgc_para_0 " compiles the exe run program file name of rear generation for the present invention; " l Changsha ", for treating the VectorLayer of rasterizing, " Changsha " represents VectorLayer title; "～data changsha.shp " is source data; "～data test_result.img " represents output file.
Step 2: adopt the mode of serial to create shapefile target data source and figure layer file, then close target data source file.Create the class that first shapefile target data source file will obtain the driver(establishment type of data format of " ESRI Shapefile " form in OGR storehouse before), CreateDataSource () method by driver creates shapefile target data source file, then by the CreateLayer () method of figure layer, creates the figure layer file of target data source; Finally close target data source file;
Step 3: open shapefile original to be processed, obtain the graph data amount of each figure layer of shapefile and each figure layer.First the OGRSFDriverRegistrar::Open () calling in OGR storehouse opens shapefile original, then use GetLayerByName () function to obtain figure layer, finally by GetFeatureCount (), obtain the graph data amount of each figure layer;
Step 4:MPI parallel initialization, obtains numbering and the quantity of process, and sets principal and subordinate's process.Call MPI_Init () the function parallel initialization of MPI function library, call MPI_Comm_rank () function and obtain process numbering, call the quantity that MPI_Comm_size () obtains process.0 process of setting is host process, and non-zero process is from process;
Step 5: carry out data division, determine the initial FID of figure each shapefile file of processing from process and stop FID according to figure sum in individual process processing graphics quantity=figure layer/(3 * (process sum-1)); Division methods to shapefile file data in step 5, former data are divided into a plurality of data sets, by a plurality of processes, cyclically process and write again a plurality of data sets, to reduce the granularity of the data of processing in cyclic process of individual process, parallel time between raising process, thus the efficiency of parallel processing improved.
Step 6: each enters shapefile data processing from process, what first complete shapefile data processing sends the complete message of data processing to host process from process, host process sends to this process to enter write operation instruction after receiving message, other process is still carried out data processing operation separately simultaneously, but can not enter write operation.When enter write operation from process write operation completes, transmission has been write information to host process, host process receive write dispatch successively again that other completes shapefile data processing after message from process, enter write operation, until other process completes oneself all data processing tasks and write operation.
3. beneficial effect
Compared to existing technology, beneficial effect of the present invention is as follows:
(1) solve OGR built-in function and write the data cover problem that shapefile file produces; When carrying out write operation to shapefile file simultaneously, a plurality of processes can produce figure covering problem in shapefile file, the time that this method staggers each process write operation by Slave Parallel method, control a plurality ofly from process asynchronous execution write operation, well solved the data cover problem that generation is write in multi-process simultaneously.
(2) improved the whole efficiency of shapefile file processing, when entering write operation from process for one, the data processing work that other process still can be Myself, until when entering the process of writing while having completed write operation, host process is dispatched another process that first arrives write operation again and is carried out write operation, thereby take full advantage of computational resource, improved the whole efficiency of shapefile file processing.
Accompanying drawing explanation
Accompanying drawing 1 is main program flow chart;
Accompanying drawing 2 is principal and subordinate's process communication figure;
Accompanying drawing 3 is shapefile vector data figure;
Accompanying drawing 4 is shapefile vector data concurrent write speed-up ratio figure;
Accompanying drawing 5 is shapefile vector data parallel output result images.
Embodiment
The test data that the present embodiment adopts the present status of land utilization investigation earth polygon (changsha.shp) in Changsha shown in accompanying drawing 3 to write as the shapefile file Slave Parallel based on OGR.The figure spot quantity of test data is 692176, and data volume size is 932M, 11819.46 square kilometres of the overlay area total areas.
The environment of the concrete operation of this example is IBM System x3500-M3X server catalyst Catalyst environment.Server hardware is configured to CPU2, (specification is Intel Xeon Quad Core E5620, dominant frequency 2.4GHz, 12MB Cache, four cores); Inside save as 8GB(2 root 4GB memory bar, specification is DDR31333MHz LP RDIMM); Hard disk is 2TB(4 500GB hard disk, and specification is 7.2K6Gbps NL SAS2.5-inch SFF Slim-HS HD), network is integrated twoport gigabit Ethernet.Software configuration: operating system is selected Centos Linux, wherein the product of realizing of MPI is selected OpenMPI, and more than version 1.4.1, GDAL selects 1.81 with adjustment of the printing plate.
Concrete implementation step is as follows:
Step 1: input command line parameter: mpirun-np8hpgc_para_O-l Changsha～data changsha.shp～data test_result.img;
Wherein, ' np ' ' l ' is the boot symbol of order line, and ' mpirun ' represents to call the statement parameter of MPI application program; ' np8 ' represents to call 8 process numbers; ' hpgc_para_O ' compiles the exe run program file name of rear generation for the present invention; ' l changsha ' is for treating the VectorLayer of rasterizing, and VectorLayer name is called changsha;～data changsha.shp be source data ,～data chansha_out.shp be output file.
Step 2:(1) mode of program serial creates Changsha shapefile target data source and figure layer file (changsha_out.shp), then closes target data source file (as shown in Figure 1).(2) create the driver that first shapefile target data source file will obtain " ESRI shapefile " form in OGR storehouse before, CreateDataSource () method by driver creates shapefile target data source file, then by the CreateLayer () method of figure layer in OGR storehouse, creates the figure layer file of target data source; (3) finally call CloseFile () function and close target data source file;
Step 3: open changsha.shp file, obtain the graph data amount of changsha.shp figure layer and figure layer.First the OGRSFDriverRegistrar::Open () calling in OGR storehouse opens changsha.shp file, then use GetLayerByName () function to obtain changsha.shp file map layer, finally by GetFeatureCount () obtain figure layer graph data amount;
Step 4: use the MPI_Init () function in MPI storehouse to carry out parallel initialization, call MPI_Comm_rank () function and obtain process numbering (rank), call the quantity (size) that MPI_Comm_size () obtains process.0 process of setting is host process (master_o ()), and non-zero process is from process (slave_o ()) (as shown in Figure 1).
Step 5: carry out data division according to the graph data quantity of the quantity from process and former changsha.shp file map layer, determine initial FID and the quantity of figure (feature) each shapefile file of processing from process, this example has been divided into 3 data sets former data, thereby reduces the granularity of the data of individual process processing; The core code that data are divided is:
Step 6: each enters from process FeatureN the feature data processing (DoWell ()) of distributing to separately, what first complete Feature data processing sends the complete message of data processing to host process (master_o ()) from process (slave_o ()), master_o () sends to this process to enter write operation instruction after receiving message, other slave_o () still carries out data processing operation separately simultaneously, but can not enter write operation.After entering the slave_o () write operation of write operation and completing, transmission has been write information to master_o (), host process is received to have write and is dispatched successively other slave_o () that completes shapefile data processing after message again and enter write operation, until other process completes oneself all data processing tasks and write operation (as Fig. 2).Final Output rusults is as Fig. 5, Fig. 5 and Fig. 3 comparison, and the data volume of Fig. 5 (figure spot quantity is 692176, and data volume size is 932M) is all identical with Fig. 3 with figure spot number, and this concurrent write method is correct as seen.Master routine core code is as follows:
1) DoWell () is shapefile file processing time delay function.Because do not relate to shapefile file data processing method in this method, but in order to consider the impact of this processing, data processing operation is served as in the operation of having selected multi-process to clone (Clone ()) to the graph data in former data.Concrete core code is as follows:
2) host process controls by the quantity of the feature of figure layer the information that host process reception sends from process, thereby reaches the object of mating with the quantity of information sending from process.Principal and subordinate's process communicates by the MPI_Send () in MPI built-in function and MPI_Recv ().
MPI_Send(buf,count,datatype,dest,tag,comm)
Buf: the start address (optional type) that sends buffer zone
Count: by the number (nonnegative integer) of the data that send
Datatype: the data type (handle) that sends data
Dest: object process identification number (integer)
Tag: message flag (integer)
Comm.: communication domain (handle)
MPI_Recv(buf,count,datatype,source,tag,comm,status)
Buf: the start address of reception buffer zone (optional data type)
Count: the number of maximum receivable data (integer)
Datatype: the data type (handle) that receives data
Source: the source that receives data sends the process identification number (integer) of data
Tag: message identifier, with the expression of corresponding transmit operation match identical (integer)
Comm.: the communication domain (handle) at this process and transmission process place
Status: return state (Status Type)
Host process core code is as follows:
3) from process, file write operation is passed through to File Open function OpenFile (), file writes function WriteFile () and closing of a file function CloseFile () completes.Concrete core code is as follows:
In order to detect the efficiency of the shapefile file Slave Parallel write method based on OGR, the method is carried out under high-performance calculation machine platform to the test of writing of changsha.shp file, this high performance platform is the mobile workstation of 4 core 8 threads.Choosing respectively 1,2,4,6,8,10,12,14,16 process tests.The efficiency of the shapefile file Slave Parallel write method based on OGR is recently evaluated in the acceleration of choosing each process by calculating.Evaluation result is in Table 1 and Fig. 4.
The efficiency rating result of table 1 concurrent write method
Efficiency rating result shows, the efficiency of the shapefile file Slave Parallel write method based on OGR is obviously improved.But the attention of value, the time speed-up ratio obtaining when process number is 2 is less than 1, this is because the character of master-slave mode determines, Slave Parallel strategy needs a host process to control from process, so when having two processes to process, in fact only have a process in the processing of carrying out data, and need to communicate with host process, spent some times.Theoretical mxm. should reach maximal value when 16 processes make full use of the computational resource of computing machine; But when 8 process, reach maximal value, this is due to middle data processing (DoWell()) process is along with process number increases, the time of individual process data processing cost is fewer and feweri, write latency and call duration time improve, make the reduction of 8 processes efficiency regularity afterwards, meet theory and actual conditions.After 16 processes, computing machine enters overload operation, and because computing machine calculates communication issue between performance and process, time speed-up ratio drops to below 1, also meets theoretical reasoning.From whole test result, the raising of the shapefile file Slave Parallel write method efficiency based on OGR is that significantly the invention of the method is successfully.
Claims (2)
1. the shapefile file Slave Parallel write method based on OGR, the steps include:
Step 1: input command line parameter: mpirun-np8hpgc_para_O-l Changsha～data changsha.shp～data test_result.img;
Wherein, " np " " l " is the boot symbol of order line, and " mpirun " represents to call the statement parameter of MPI application program; " n p8 " represents to call 8 process numbers; " hpgc_para_0 " compiles the exe run program file name of rear generation for the present invention; " l Changsha ", for treating the VectorLayer of rasterizing, " Changsha " represents VectorLayer title; "～data changsha.shp " is source data; "～data test_result.img " represents output file;
Step 2: the mode of serial creates shapefile target data source and figure layer file, then closes target data source file; Create the driver that first shapefile target data source file will obtain " ESRI Shapefile " form in OGR storehouse before, CreateDataSource () method by driver creates shapefile target data source file, then by the CreateLayer () method of figure layer, creates the figure layer file of target data source; Finally close target data source file;
Step 3: open shapefile original to be processed, obtain each figure layer of shapefile and the graph data amount of each figure layer; First the OGRSFDriverRegistrar::Open () calling in OGR storehouse opens shapefile original, then use GetLayerByName () function to obtain figure layer, finally by GetFeatureCount (), obtain the graph data amount of each figure layer;
Step 4:MPI parallel initialization, obtains numbering and the quantity of process, and sets principal and subordinate's process; Call MPI_Init () the function parallel initialization of MPI function library, call MPI_Comm_rank () function and obtain process numbering, call the quantity that MPI_Comm_size () obtains process; 0 process of setting is host process, and non-zero process is from process;
Step 5: carry out data division, determine the initial FID of figure each shapefile file of processing from process and stop FID according to figure sum in individual process processing graphics quantity=figure layer/(3 * (process sum-1));
Step 6: each enters shapefile data processing from process, what first complete shapefile data processing sends the complete message of data processing to host process from process, host process sends to this process to enter write operation instruction after receiving message, other process is still carried out data processing operation separately simultaneously, but can not enter write operation; When enter write operation from process write operation completes, transmission has been write information to host process, host process receive write dispatch successively again that other completes shapefile data processing after message from process, enter write operation, until other process completes oneself all data processing tasks and write operation.
2. the shapefile file Slave Parallel write method based on OGR according to claim 1, it is characterized in that the division methods to shapefile file data in step 5, former data are divided into a plurality of data sets, by a plurality of processes, cyclically process and write again a plurality of data sets, to reduce the granularity of the data of processing in cyclic process of individual process, parallel time between raising process, thus the efficiency of parallel processing improved.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110441737.3A CN102591709B (en) | 2011-12-20 | 2011-12-27 | Shapefile master-slave type parallel writing method based on OGR (open geospatial rule) |
Applications Claiming Priority (3)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110429888 | 2011-12-20 | ||
| CN201110429888.7 | 2011-12-20 | ||
| CN201110441737.3A CN102591709B (en) | 2011-12-20 | 2011-12-27 | Shapefile master-slave type parallel writing method based on OGR (open geospatial rule) |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102591709A CN102591709A (en) | 2012-07-18 |
| CN102591709B true CN102591709B (en) | 2014-03-12 |
Family
ID=46480417
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201110441737.3A Expired - Fee Related CN102591709B (en) | 2011-12-20 | 2011-12-27 | Shapefile master-slave type parallel writing method based on OGR (open geospatial rule) |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102591709B (en) |
Families Citing this family (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103164267A (en) * | 2013-03-29 | 2013-06-19 | 汉柏科技有限公司 | Lock-free message queue implementation method |
| CN103678705B (en) * | 2013-12-30 | 2016-07-13 | 南京大学 | A kind of VCT file is to the vector data Parallel transformation method of shapefile file |
| CN103927374B (en) * | 2014-04-24 | 2015-09-09 | 盐城工学院 | A kind of Web GIS map layer creation method based on SharpMap |
| CN105975331A (en) * | 2016-04-26 | 2016-09-28 | 浪潮(北京)电子信息产业有限公司 | Data parallel processing method and apparatus |
| CN106373119A (en) * | 2016-09-05 | 2017-02-01 | 广东工业大学 | Fiber detection method and system |
| CN106611037A (en) * | 2016-09-12 | 2017-05-03 | 星环信息科技（上海）有限公司 | Method and device for distributed diagram calculation |
| CN111061652B (en) * | 2019-12-18 | 2021-12-31 | 中山大学 | Nonvolatile memory management method and system based on MPI-IO middleware |
| CN111915281B (en) * | 2020-09-03 | 2021-02-23 | 深圳市不动产评估中心(深圳市地质环境监测中心) | Block evaluation approval method and system based on geological disaster easily-issued subarea, intelligent terminal and storage medium |
| CN112463905B (en) * | 2020-11-30 | 2022-06-03 | 湖北金拓维信息技术有限公司 | Vector data SHP file parallel writing method |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20010037981A (en) * | 1999-10-21 | 2001-05-15 | 이계철 | Method for converting map data to data for object_related geographic information system |
| CN101799762A (en) * | 2010-04-07 | 2010-08-11 | 中国科学院对地观测与数字地球科学中心 | Quick parallelization programming template method for remote sensing image processing algorithm |
| CN102200575A (en) * | 2010-12-02 | 2011-09-28 | 南京大学 | Image ship detection method based on constant false alarm rate |
-
2011
- 2011-12-27 CN CN201110441737.3A patent/CN102591709B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20010037981A (en) * | 1999-10-21 | 2001-05-15 | 이계철 | Method for converting map data to data for object_related geographic information system |
| CN101799762A (en) * | 2010-04-07 | 2010-08-11 | 中国科学院对地观测与数字地球科学中心 | Quick parallelization programming template method for remote sensing image processing algorithm |
| CN102200575A (en) * | 2010-12-02 | 2011-09-28 | 南京大学 | Image ship detection method based on constant false alarm rate |
Non-Patent Citations (1)
| Title |
|---|
| 孙宁，蒲英霞.使用Python函数式编程进行空间回归分析.《地理信息世界》.2009,(第3期), * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102591709A (en) | 2012-07-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102591709B (en) | Shapefile master-slave type parallel writing method based on OGR (open geospatial rule) | |
| Zhang et al. | Enabling in-situ execution of coupled scientific workflow on multi-core platform | |
| Ellsworth et al. | Dynamic power sharing for higher job throughput | |
| CN103207774B (en) | For solving the method and system of thread divergence | |
| CN104536937B (en) | Big data all-in-one machine realization method based on CPU GPU isomeric groups | |
| CN103744716B (en) | A kind of dynamically interruption Well-Balanced Mapping method based on current VCPU dispatch state | |
| CN102279766B (en) | Method and system for concurrently simulating processors and scheduler | |
| CN103761215B (en) | Matrix transpose optimization method based on graphic process unit | |
| CN103440163A (en) | PIC (Peripheral Interface Controller)-model-based accelerator simulation method implemented by using GPU (Graphic Processing Unit) in parallel | |
| CN103309786A (en) | Methods and apparatus for interactive debugging on a non-pre-emptible graphics processing unit | |
| CN103425534A (en) | Graphics processing unit sharing between many applications | |
| CN107370835A (en) | A kind of cloud computing center network architecture based on SDN and NFV technologies | |
| CN106648552A (en) | CPU/GPU synchronization mechanism | |
| CN104267940A (en) | Quick map tile generation method based on CPU+GPU | |
| Montella et al. | Virtualizing high-end GPGPUs on ARM clusters for the next generation of high performance cloud computing | |
| CN103996216A (en) | Power efficient attribute handling for tessellation and geometry shaders | |
| CN103262035A (en) | Device discovery and topology reporting in a combined CPU/GPU architecture system | |
| Wu et al. | Using hybrid MPI and OpenMP programming to optimize communications in parallel loop self-scheduling schemes for multicore PC clusters | |
| CN106484532B (en) | GPGPU parallel calculating method towards SPH fluid simulation | |
| Liang et al. | A Survey on Spatio-temporal Big Data Analytics Ecosystem: Resource Management, Processing Platform, and Applications | |
| Khaitan et al. | Proactive task scheduling and stealing in master-slave based load balancing for parallel contingency analysis | |
| Han et al. | Distributed loop scheduling schemes for cloud systems | |
| Liu et al. | BSPCloud: A hybrid distributed-memory and shared-memory programming model | |
| Parashar | Big data challenges in simulation-based science. | |
| Weisz et al. | Graphgen for coram: Graph computation on FPGAs |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20140312 Termination date: 20171227 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |