Answer
In FME 2011, we have four formats that are either exclusively point cloud formats (LAS and POD), or can be interpreted as a point cloud or a set of individual points (XYZ). We also support them along with other geometries in our own FFS format.
Formats
LAS
LAS format is defined in it specification as follows:
The LAS file is intended to contain LIDAR point data records.The data will generally be put into this format from software(e.g. provided by LIDAR hardware vendors) which combines GPS,IMU, and laser pulse range data to produce X, Y, and Z pointdata. The intention of the data format is to provide an openformat that allows different LIDAR hardware and software toolsto output data in a common format.
LAS is the most popular point cloud data - most available datasets are stored in LAS format (with XYZ following). Often, the word LAS is used as a synonym to LiDAR, which is not quite correct but reflects the popularity of the format.
FME versions before FME 2011 were reading LAS files (up to version 1.1) as sets of individual points, which was very slow and inefficient in further processing.
With FME 2011, we are able to read and write the whole LAS file (up to version 1.2) as a single point cloud geometry. This increases reading and writing speed from minutes and hours to seconds.
Note that for compatibility reasons, the workspaces created with older versions of FME will still read LAS files as individual points (fme_point). This means that if you replace your reader with the new one, you also will have to review the whole transformation workflow. However, performance gain is so significant that it really makes sense to make such a change.
POD
POD (Point Database) is a
Pointools (high performance point cloud software) native point cloud format.
FFS
FFS is a native FME format that can store anything that features can carry. Once FME got point cloud geometry support, FFS format had to be modified to accommodate these changes.
XYZ
Point cloud data often comes in a form of text files. The problem is, there is no standard definition how these files should look like, and, as result, we can see all kinds of formatting - different separators, projections, precision, different components or no components at all. Here are a few examples of XYZ data:
Example 1:x,y,z,gpstime,intensity,classification,flight_line493509.88,4245673.18,-32.00,160025.954000,0,2,1493511.98,4245674.49,-31.86,160025.954000,0,1,1493512.56,4245674.59,-31.92,160025.956800,0,1,1Example 2:GPSTime X Y Z Intensity ReturnNum NumReturns Class ScanDirection ScanAngleRank GPSWeek FlightLine151539.687767 1145355.12 549174.79 182.19 134 1 1 2 0 4 1519 154151539.687807 1145365.72 549174.33 181.99 1 2 2 2 0 3 1519 154151539.687818 1145368.41 549174.23 181.89 3 3 3 2 0 3 1519 154Example 3:589375.00 232939.90 749.60589375.01 232881.08 758.41589375.01 232901.53 758.73
XYZ Point Cloud reader allows flexibly set up reading such files, whereas XYZ writer gives users control over coordinate and components order, component selection, and a choice of separators.
Future Formats
In the next FME releases we plan to add point cloud support to the following formats:
If you know a point cloud format, which you think we should support, please send an e-mail to
support@safe.com.