Portal? Wheel column ?? Blog. csdn. netleverage_12291 Development History 2 processing process 3 storage model separates the logical structure from the physical structure (1) logical structure: these tables contain columns of the SDO_PC type that contain general attributes and the number of units in occupied space, and contain a pointer to the point cloud block table (2
Portal? Wheel column? ? Http://blog.csdn.net/leverage_1229 1 development process 2 processing process 3 storage model separating the logical structure from the physical structure (1) logical structure: these tables contain columns of the SDO_PC type that contain general attributes and the number of units in occupied space, and contain a pointer to the point cloud block table (2
Portal? Wheel column? ? Http://blog.csdn.net/leverage_1229
1. Development History
2. handling process
3. The storage model separates the logical structure from the physical structure.
(1) logical structure:
These tables contain columns of the SDO_PC type.
Contains general attributes and the number of units in Occupied Space
It also contains a pointer to a cloud block table.
(2) physical structure:
"Partition table"
Included cloud Blocks
It can be very huge.
Define the structure in the SDO_PC_BLK object type
4. There is no limit on the number of cloud tables for creating a vertex cloud table.
Supports combination of fields of any type
Scenarios can be searched by any attribute
And the scenario has a spatial margin
Create table LIDAR_SCENES (
SCENE_ID number primary key,
COLLECTION_TS TIMESTAMP,
Description clob,
... (Any number of attributes )...
POINT_CLOUD SDO_PC
);
The 5SDO_PC PC_EXTENT is the unit of space occupied by point clouds.
A spatial index is required to support spatial retrieval.
Attribute |
Description |
BASE_TABLE |
Table Name that contains a column of the SDO_PC type |
BASE_TABLE_COL |
SDO_PC column name |
PC_ID |
Point Cloud ID |
BLK_ID |
Table Name that contains the block information of the vertex cloud |
PTN_PARAMS |
Cloud partition Parameters |
PC_EXTENT |
The SDO_GEOMETRY object displays the spatial margin of the point cloud (the minimum boundary object is a closed point. All objects in the cloud) |
PC_TOL |
The value of the specified point in a point cloud. |
PC_TOT_DIMENSIONS |
The total number of dimensions in the cloud. Including spatial and other non-spatial dimensions. The maximum value is 9. |
PC_DOMAIN |
(Not used currently) |
PC_VAL_ATTR_TABLES |
The SDO_STRING_ARRAY object specifies the name of any value attribute table for the point cloud. |
PC_OTHER_ATTRS |
The XMLTYPE object specifies any other attribute of the point cloud. |
6 create a point cloud partition table using the SDO_PC_BLK type
Define a primary key on the block ID
Use SECUREFILE lobs (New 11g structure)
Allow LOBs compression (repeated encryption and deletion)
Create table PC_BLK_01 OF SDO_PC_BLK (
Primary key (
OBJ_ID, BLK_ID
)
)
LOB (POINTS) STORE AS SECUREFILE
(Compress high nocache nologging );
One segment of the 7SDO_PC_BLK type description point
Unique identifier of a block
Scenario ID (consistent with PC_ID) and block ID (BLK_ID)
Attribute |
Description |
OBJ_ID |
Point Cloud Object ID |
BLK_ID |
Block ID |
BLK_EXTENT |
Block Space Range |
BLK_DOMAIN |
(Not used currently) |
PCBLK_MIN_RES |
Minimum resolution of visible blocks in a query |
PCBLK_MAX_RES |
Maximum resolution of visible blocks in a query |
NUM_POINTS |
The total number of POINTS in BLOB. |
NUM_UNSORTED_POINTS |
Number of unordered POINTS in POINTS BLOB |
PT_SORT_DIM |
Dimension of an ordered point |
POINTS |
BLOB points |
8 BLOB structure BLOB contains an array of storage points
Point Encoding
D 64-bit floating point number (d = the dimension of the vertex)
A 32-bit integer represents the number of points
A 32-bit integer represents the number of partitions.
Future: compression format
Store coordinates from the source of the block MBR as offset
Use a short integer
9 initialize a point cloud (1) define the structure and organization of the point cloud
Resolution, dimension, and amplitude
Block capacity
(2) Specify the block location for each point cloud
Vertex block table name
Unique table ID
Insert into LIDAR_SCENES (
SCENE_ID, POINT_CLOUD)
VALUES (
1,
SDO_PC_PKG.INIT (
BASETABLE => 'lidar _ SCENES ',
BASECOL => 'point _ cloud ',
BLKTABLE => 'pc _ BLK_01 ',
PTN_PARAMS => 'blk _ CAPACITY = 1000 ',
PC_TOL = & gt; 0.005,
PC_TOT_DIMENSIONS => 3,
PC_EXTENT =>
Sdo_geometry( 2003,432 6, NULL,
SDO_ELEM_INFO_ARRAY (1, 1003, 3 ),
SDO_ORDINATE_ARRAY (-74, 40,-73, 41)
)
)
);
10 load a point cloud (1) load a point cloud from a normal table
DECLARE
PC SDO_PC;
BEGIN
SELECT POINT_CLOUD INTO PC
FROM LIDAR_SCENES WHERE SCENE_ID = 1;
SDO_PC_PKG.CREATE_PC (PC, 'input _ POINTS ');
END;
(2) Structure of the table used for Input
(3) The input table can be a text file and is associated with an external table.