|Home | Docs | Bugs | FAQ | Download|
This basic tutorial explains how to use libLAS to read and write LIDAR data encoded in LAS file format from Python.
1. Reading LAS data in Python is a simple as opening the file with the file.File class and using the iterator to chug through the points.
>>> from liblas import file >>> f = file.File('file.las',mode='r') >>> for p in f: ... print 'X,Y,Z: ', p.x, p.y, p.z
>>> from liblas import file >>> f = file.File('file.las', mode='r') >>> p = f.read(0) >>> p <liblas.point.Point object at 0x7377f0>
The LAS format, as of this writing, provides three different file format versions (1.0, 1.1, and 1.2) and four different point format verions (0, 1, 2, 3). It is important that you be mindful of the minor_version and the dataformat_id when working with files. If the dataformat_id is not correct for the type of data you want to store (color, time, or color + time), none of that data will be read or written, even though placeholders will exist on the liblas.point for them.
You control whether or not color, time, or color + time is stored in the file by setting the dataformat_id in the header that you use to create the file.File. This has to be done when you instantiate the file, it cannot be changed after the fact. Attempting to do so may result in disastrous results.
Headers, points, VLRs, colors, and GUIDs are copies, not references in the libLAS Python bindings. After opening a LAS file, you can fetch the header with the following property:
>>> header = f.header >>> h <liblas.header.Header object at 0x737790>
There are many properties of the header that you can get and set:
>>> h.major_version, h.minor_version (1, 0) >>> h.dataformat_id 0 >>> h.min [289020.90000000002, 4320942.6100000003, 166.78] >>> h.max [290106.02000000002, 4323641.5700000003, 215.48000000000002] >>> h.scale [0.01, 0.01, 0.01] >>> h.offset [-0.0, -0.0, -0.0] >>> h.point_records_count 3265110L >>> from liblas import guid >>> h.project_id '00000000-0000-0000-0000-000000000000' >>> h.guid 00000000-0000-0000-0000-000000000000 >>> g = guid.GUID() >>> g 5cb59173-124b-476b-9729-bafa87cfb27c >>> h.guid = g >>> h.guid 5cb59173-124b-476b-9729-bafa87cfb27c >>> h.project_id '5cb59173-124b-476b-9729-bafa87cfb27c' # only works if you have libgeotiff and/or GDAL linked in >>> s = h.srs >>> s.proj4 +proj=tmerc +lat_0=0.000000000 +lon_0=-93.000000000 +k=0.999600 \ +x_0=500000.000 +y_0=0.000 +ellps=WGS84 +units=m
The liblas.point module contains a Point class that you can use to manipulate LAS point data. It is fairly basic and contains a number of properties you can set and get:
>>> p.x, p.y, p.z (289814.15000000002, 4320978.6100000003, 170.75999999999999) >>> p.scan_angle 0 >>> p.scan_direction 0 >>> p.return_number 0 >>> p.number_of_returns 6 >>> p.flightline_edge 0 >>> p.classification 2 >>> p.time datetime.datetime(1970, 1, 6, 12, 44, 10, 1) >>> p.intensity 120 >>> c = p.color >>> c.red 255 >>> c.blue 255 >>> c.green 255
Variable Length Records (VLR) are frequently used by applications to store anything they wish in the file as a “blob” written into the header of the file. libLAS supports writing and creating your own VLRs in addition to taking on the work of interpreting and using VLR records related to spatial reference systems if GDAL and proj.4 are linked into the library.
The following code demonstrates how to write your own VLR by opening an XML file and inserting it into a new file.
from liblas import file as lasfile from liblas import vlr from liblas import header as lasheader f = lasfile.File('test/data/srs_utm17.las',None,'rb') h = f.header v = vlr.VLR() text = open('schemas/las.xml','rb').read() import ctypes data = ctypes.create_string_buffer(text) v.userid='hobu' v.recordid = 12345 v.data = data h.add_vlr(v) f2 = lasfile.File('junk.las',header=h,mode='w') for p in f: f2.write(p) f2.close()
To write a new LAS file, you are first required to have a header. The header will have a number of default values, but it is important to set the dataformat_id and version_minor if you wish to have 1.1 files or records that also include time values.
>>> from liblas import header >>> h = header.Header() ### Support storing time values >>> h.dataformat_id = 1 ### Store a 1.1 version file >>> h.minor_version = 1
Another important item to not is possible to have the same file open for read and write at the same time because LAS files are sequential. For example, the following will fail:
>>> f = file.File('junk.las', mode="w", header=h) >>> f2 = file.File('junk.las') Traceback (most recent call last): ... LASException: ('File %s is already open. Close the file or delete the reference to it', 'junk.las')
Writing to a LAS file is as simple as opening the file for write mode with the header you want to write and issuing the write() command with some liblas.point.Point instances:
>>> f = file.File('junk.las',mode='w', header= h) >>> pt = liblas.point.Point() >>> f.write(pt) >>> f.close()