PointCloudIO_ptx Class Reference

Implementation of file input/output operations for ASCII Cyclone point cloud export format (PTX). More...

#include <easy3d/fileio/point_cloud_io_ptx.h>

Public Member Functions
	PointCloudIO_ptx (const std::string &file_name)
	TODO: Translator not implemented.
PointCloud *	load_next ()
	Reads a single point cloud from the file.

Detailed Description

Implementation of file input/output operations for ASCII Cyclone point cloud export format (PTX).

PTX is an ASCII based interchange format for point cloud data. It utilizes the concept of separate scans, each with points defined in their own coordinate system and a 'Registration' of all of those point clouds into a single coordinate system. The point data for each cloud is stored in its original coordinate system and a matrix of transforms for each point cloud are provided as header information.

A PTX file can have one or more point cloud. Each point cloud starts with a header. Each following line then describes one point in the cloud. Cyclone exports PTX with 7 columns when the cloud has RGB values from the digital camera (x, y, z, intensity, red, green, blue). Red, Green, Blue have the integer range [0, 255]. A cloud in PTX has 4 columns (x, y, z, intensity) when the cloud does not have RGB values. PTX intensities use the decimal range [0, 1]. Individual values on the same line are separated by a blank space. The coordinate unit is always in meters. PTX can only be used on "[[KB:132:gridded]]" clouds (directly from scans), not unordered or unified clouds. Each cloud is "fully populated", in that even missing points are represented (as a line with "0 0 0" for the XYZ coordinate). This gives the importing software enough information to restore the original scan-line ordering of the cloud (for example, to estimate normal vectors).

PTX point cloud header:

number of columns
number of rows
st1 st2 st3 ; scanner registered position
sx1 sx2 sx3 ; scanner registered axis 'X'
sy1 sy2 sy3 ; scanner registered axis 'Y'
sz1 sz2 sz3 ; scanner registered axis 'Z'
r11 r12 r13 0 ; transformation matrix
r21 r22 r23 0 ; this is a simple rotation and translation 4x4 matrix
r31 r32 r33 0 ; just apply to each point to get the transformed coordinate
tr1 tr2 tr3 1 ; use double-precision variables

The first four lines of three numbers each are the position and primary axes of the scanner after any registration/transformation. The next four lines of four numbers each may look similar in some cases, but if you have a non-identity UCS when the PTX was exported, the numbers will look different. If the cloud was untransformed by a registration (or not registered), the first four lines of three numbers each would be 0,0,0; 1,0,0; 0,1,0; 0,0,1. The 4x4 matrix may not be identity if there is a UCS applied.

Note that a cloud that was registered may have a 4x4 identity matrix due to a UCS that is set to that scanner's registered position. For both formats, in a point cloud with color (i.e., the R G B values a re present for each point in that cloud), the RGB value (0, 0, 0) is reserved to mean "no color". An example where this might happen would be if a digital camera did not cover that point. That point then has "no color", even though other points may have color.

Example data file:

4.148210 12.076586 1.381819 0.461036 89 87 112
4.147804 12.075443 1.382720 0.458854 89 87 112
4.148891 12.078566 1.384060 0.461967 87 86 111   End coordinate list (cloud #1)
40                                               Scan 40 x 40 degree (cloud #2)
40
4.151261 1.736411 -8.110131                      Begin transformation (cloud #2)
-0.434947 -0.017233 -0.900291
-0.240839 0.965618 0.097871
0.867650 0.259394 -0.424143
-0.434947 -0.017233 -0.900291 0
-0.240839 0.965618 0.097871 0
0.867650 0.259394 -0.424143 0
4.151261 1.736411 -8.110131 1
4.146983 12.074142 1.134998 0.488502 76 74 91     Begin coordinate list (cloud #2)
4.147541 12.075727 1.136227 0.486915 76 74 91     Format: X Y Z intensity R G B
4.147919 12.076833 1.137371 0.486976 76 74 91     Intensity range: 0 - 1
4.148849 12.079544 1.138655 0.485328 76 74 91
4.146931 12.073991 1.139137 0.486076 76 74 91
4.148440 12.078399 1.140532 0.488441 79 76 93

Note

• PTX files hold the intensities in a range 0-1
• PTS files are -2048 to 2047 as integers (no fractions)
• If you need to convert PTS to PTX intensity values you need to add 2048 and divide by 4096.

See http://w3.leica-geosystems.com/kb/?guid=5532D590-114C-43CD-A55F-FE79E5937CB2

Example usage:

PointCloudIO_ptx serializer(file_name);
PointCloud* model = nullptr;
while (model = serializer.load_next()) {
    addModel(model);
}

---

The documentation for this class was generated from the following files:

• /Users/lnan/Documents/Projects/Easy3D/easy3d/fileio/point_cloud_io_ptx.h
• /Users/lnan/Documents/Projects/Easy3D/easy3d/fileio/point_cloud_io_ptx.cpp