HET51_WWO_TLS
The HET51_WWO_TLS scene has been obtained from the Wytham Woods 3D model which represents a one hectare deciduous forest in Wytham, UK. Calders et al. (2018) matched the TLS data with traditional census data to determine the species of each individual tree and allocate species-specific radiometric properties.
The 3D structure of the canopy and of each modelled tree are stored in ASCII obj files (a modified version of the open Wavefront format https://www.fileformat.info/format/wavefrontobj/egff.htm).
Whether the original Wytham Wood model features a sloping terrain model, in RAMI-V all trees have been shifted along the z-axes to lay on a common flat surface. This guarantees continuity on the boundaries and energy-conservation in an infinite replication scheme. Additionally, the scene have been cropped to fit a 100x100m bounding-box (Table 2). To do that, any primitive (triangle, cylinder or disc) laying outside the new bbox has been omitted or commented from the original files.
Figures 1 and 2 show a graphical rendering of the HET51_WWO_TLS scene as obtained with Rayshade program.
The scene definition consists of three main parts:
Note: lm_1638 was removed in this release of the scene as it does not contain primitives within the selected bounding box.
All these files are stored in the zipped archives linked in Table 1.
The leaves are represented with a triangular meshing. The Wavefront files are compliant, for leaves only, with the import tool of Blender software (suggest v2.80+), which allows to export the meshing in several additional formats. As the stems are represented with groups of cylinders, and the standard Wavefront format does not foresee a keyword to represent such a primitive, the Blender importer will not manage the original ww_*.obj files.
The Rayshade like files (.def) have been converted with a Python tool developed at the Joint Research Centre and contain the definition of triangles position and orientation in the scene. This format might be readable with respect to the Wavefront representation. Users should adopt the one more convenient to produce a suitable input for their code.
Terrain should be defined as a flat surface at z=0 with horizontal coordinates (x,y) from (40,100) to (140,200).
The overall scene is created by distributing the trees (stem and crown) into a unique scenario. This is done group named scene_part_all defined in the file scene_all.obj(def). The group results itself from the composition of 133 groups named scene_part_i (files scene_part_i.obj(def), with i varying from 1 to 143 (note: there are missing undefined parts in the series: 3, 32, 37, 50, 51, 61, 63, 65, 74, 81).
Elements | Wavefront * | Rayshade ** | Num files |
---|---|---|---|
Stems | obj | def | 558 |
Leaves | obj | def | 528 |
Scene definition | obj | def | 1 |
The stems and leaves, minimum and maximum coordinates are given, in meters, in Table 2.
X (m) | Y (m) | Z (m) | ||||
---|---|---|---|---|---|---|
From | To | From | To | From | To | |
Stems | 40.0 | 140.0 | 100.0 | 200.0 | 0.0 | 30.5 |
Leaves | 40.0 | 140.0 | 100.0 | 200.0 | 0.0 | 30.5 |
The center of the scene, with respect to which all measurements should refer, is set arbitrarily to
($x_c$, $y_c$) = (90 m, 150 m).
The Wytham Wood scene is described by using triangles and cylinders primitives.
For what concerns the Wavefront .obj files the triangle representation follows the standard format as described in https://www.fileformat.info/format/wavefrontobj/egff.htm.
usemtl lm_ACERPS v 96.09 164.13 6.81 v 96.13 164.00 6.65 v 96.03 163.99 6.57 f -3 -2 -1
Alternatively, the rayshade format represents triangle through the following statement
triangle Tag(lm_ACERPS) lm_ACERPS 96.09 164.13 6.81 96.13 164.00 6.65 96.03 163.99 6.57
In the extended version of the wavefront files (ww_*.obj) the keyword ccyl is used to define closed cylinders, such that rays can't get inside the trunk at the joints. In the snippets below: 1) "g box 16250" is the bounding box definition and will be inside some pair of "{}". 2) "usemtl ww_ACERPS" defines the material to use (defined in the spectral properties file below); 3) the two lines beginning with "v" define some vertices (here they are used to define the center of the two cylinder ends, so that the cylinder will be around the line drawn with those two points; 4) "ccyl -2 -1 0.139679190346" as mentioned is the definition of a closed cylinder with two ends at the vertices defined in the previous two lines (-1 and -2 lines), and the final number is its radius (0.14).
. g_box 16250 usemtl ww_ACERPS v 92.7 162.0 -11.6 v 92.4 162.2 -11.8 ccyl -2 -1 0.14 .
Similarly, in the Rayshade format (ww_*.def) files, cylinders are defined as follows
. name box_16250 list cylinder Tag(ww_ACERPS) ww_ACERPS 0.14 92.4 162.2 -11.8 92.7 162.0 -11.6 disc Tag(ww_ACERPS) ww_ACERPS 0.14 92.4 162.2 -11.8 -0.002975 0.001314 -0.001998 disc Tag(ww_ACERPS) ww_ACERPS 0.14 92.7 162.0 -11.6 0.002975 -0.001314 0.001998 end .
where
In v2 HET51 files, only the statements
starting with 'f' and 'ccyl' (instantiation of facets and closed cylinders, respectively)
have been treated by commenting them '#' from the original .obj files.
Be aware that, all vertices ('v'), including those laying outside the new bounding box,
have been kept in the file as uncommented vertices though they are not used, as in the following example:
v 39.9981701856 103.529603496 17.499 v 40.0710467167 103.589127267 17.317 #ccyl -2 -1 0.0404831914863 v 39.378400 100.452300 14.2442 v 39.527700 100.361600 14.3012 v 39.504800 100.253800 14.3299 #f -3 -2 -1
The following code can used to comment unwanted vertices in v2 .obj files:
for f in lm/lm_*.obj do awk '$0 !~ /#f/ { line[NR] = $0 } $0 ~ /#f/ { line[NR] = $0 ; for (i=(NR-1);i>=(NR-3);i--) line[i] = "#" line[i] } END{ for (i=1;i<=NR;i++) print line[i]}' $f > $f.fullcommented done for f in ww/ww_*.obj do awk '$0 !~ /#ccyl/ { line[NR] = $0 } $0 ~ /#f/ { line[NR] = $0 ; for (i=(NR-1);i>=(NR-2);i--) line[i] = "#" line[i] } END{ for (i=1;i<=NR;i++) print line[i]}' $f > $f.fullcommented done
All of the foliage Wood and background components in the Wytham Woods scene feature LAMBERTIAN scattering properties. The tables below contain the magnitudes of the reflectance and transmission characteristics of the various canopy components for fourteen different spectral bands. The experimental identifier for the Wytham Wood scene is given by HET51_WWO_TLS_BBB_zZZaAAA where BBB is one of the spectral bands of RAMI-V (O03,O04,O06,O08,O10,O11,O12,M08,O17,MD5,M11,MD7,M12,GED). An ASCII (text) file summarising all of this information can be found here.
Trunks and branches have the same reflectance within a same tree species, and only branches is shown in the spectral properties' file and the graph below.
The illumination conditions are very likely dependent on the kind of measurement in RAMI-V more than in previous RAMI phases. For brf*, dhr, fabs*, ftran* measurements, except brf_sat, the illumination were listed in the description of measure brfpp, and duplicated in other measure description pages. For these geometries the tag will be _zZZaAAA_ with ZZ and AAA defining $\theta_i$ and $\phi_i$, respectively. In addition, diffuse isotropic illumination is foreseen for bhr, fabs*, ftran* measures (geometry tag will then be _DIFFUSE_). lidar* like measurements and thp illumination are described in the relevant measure description pages, and are the same for all scenes for which they are foreseen.
Scene | Site | Jan | Apr | Jul |
---|---|---|---|---|
HET51_WWO_TLS | Wytham Wood | _z35a138_ | _z46a147_ | _z75a154_ |