Data samples
Espoo maple was measured in Norway in leaf-off conditions and the system consisted of a phase-shift based terrestrial laser scanner, namely the Leica HDS6100 with a 650-690 nm wavelength. The distance measurement accuracy and the point separation angle of the scanner were about 2-3 mm and 0.036 degrees, respectively. The horizontal distance of the scanner to the trunk was about 7-12 m, giving an average point density on the surface of the trunk (at the level of the scanner) for a single scan around 2-5 points per square centimeter. The QSM of the subject maple tree consists of 19,000 cylinders approximating 3,078 branches.
The data sets for this QSM: Branching (EspooBraData.txt) and Segment (EspooSegData.txt) features. One can import these features into Matlab using the followingBayesForest commands:
espoo = import_qsm_data('EspooBraData.txt','EspooSegData.txt','tmp.mat');
[bra, seg, qsm_tree] = gen_scatter2('tmp.mat');The first command returns a tree object espoo (similarly, the second command
returns the same object as qsm_tree), which can be drawn as:
figure;
espoo.draw;Additionally, import_qsm_data() creates the file tmp.mat in the current
directory, which is subsequently processed by gen_scatter2() to return Matlab
objects containing scatters of this QSM, i.e. bra and seg. The file tmp.mat can
be deleted later by delete('tmp.mat'); or by any other means.
This plot of pine trees was measured near Ruotsinkylä, Finland. The plot contains six
pine trees. The structural characteristics of the QSM reconstruction were saved in
.mat files and can be accessed directly via gen_scatter2 command:
[bra, seg, qsm_tree] = gen_scatter2('qsm_X.mat');where qsm_X.mat is one of the following:
- Pine 1 (direct link)
- Pine 2 (direct link)
- Pine 3 (direct link)
- Pine 4 (direct link)
- Pine 5 (direct link)
- Pine 6 (direct link)
For example, for pines 5, 3, and 1 after loading the data with gen_scatter2,
qsm_tree.draw for each of the pine trees on the same figure will produce the
following figure:
