Provided with a point cloud, it simulates a photo as if taken from a camera in the point cloud. A text file is used to provide camera locations and orientations, as well as other per-camera parameters. How much light arrives to the lens is calculated by projecting points on a grid with 5 lens projection models: equiangular, equidistant, stereographic, orthographic and rectilinear(standard perspective).

It can be used for several purposes, ranging the fields of ecology/forestry, urban light pollution, etc...

See ouput examples at the bottom of this page .

Download the compiled executable lasPhotoCamSIM (linux) or lasPhotoCamSIM.exe (Windows 64-bit compiled with MingW) and enjoy.

To compile from source code clone this directory or download the CPP and HPP files and follow these instructions.

• Download LAStools (https://rapidlasso.com/lastools/)[https://rapidlasso.com/lastools/].
• Compile LAStools/LASlib :
  • go to LAStools directory and create a "build" directory (e.g. mkdir build)
  • cd build
  • cmake ../ (install cmake if you don't have it)
  • make
• Compile lasPhotoCamSIM :
  • go to the "example" directory in LASlib (/LASlib/example) and copy/move the lasPhotoCamSIM.cpp and lasPhotoCamSIM.hpp files there.
  • Open "Makefile" file and modify contents:

convert

all: lasexample lasexample_write_only lasexample_add_rgb lasexample_simple_classification lasexample_write_only_full_waveform lasexample_write_only_with_extra_bytes

*to *

all: lasexample lasPhotoCamSIM lasexample_write_only lasexample_add_rgb lasexample_simple_classification lasexample_write_only_full_waveform lasexample_write_only_with_extra_bytes

and right after add:

lasPhotoCamSIM: lasPhotoCamSIM.o
${LINKER} ${BITS} ${COPTS} lasPhotoCamSIM.o -llas -o $@ ${LIBS} ${LASLIBS} $(INCLUDE) $(LASINCLUDE)

You should be able then to run successfully the command "make lasPhotoCamSIM" in the directory and this creates the executable.

Use the MingW compiler chain and follow the steps above, like for linux platforms.

This tools basically estimates how much light arrives at a certain spot/plot., or how much open skype (aka gap-fraction in forests), or complementary canopy cover ratio, is present at a certain spot. User provides a CSV file with a list of coordinates that correspond to camera positions, height of camera position, and the tool reprojects the points from the point cloud to spherical coordinates with respect to a hemispherical dome around the camera, and figures how much obstruction they create, by testing 1° azimuth x 0.5° zenith sectors, rays that are traced between camera center and points.

-loc <file path>: is the path to a CSV file - with header - with the following information:

• X Y and Z: coordinates of camera locations (mandatory);
• pitch, yaw, roll: orientation of camera; (see -ori for more info.) This overrides the -ori command-line parameters if present.
• proj: also overrides the -proj parameter (see -proj for more info.). I.e. you can set for each camera the lens projections to use.
• orast: also overrides the -orast parameter (see -orast for more info.) - each camera can project to a specific size of grid - for now only square grids are supported.
• other columns can be present and will be saved in output. Comma, tab, pipe, space, column and semi-column characters are accepted as column separators. If you don't care about camera Z coordinate (e.g. if your cloud is normalized to ground) and want a fixed value, you can put '0' for the third column and fix the value using -zCam

Example 1: -loc cameras.csv cameras.csv with the following contents:

X|Y|Z
279890|5718602|0
279955|5718681|0
279880|5718759|0
279963|5718737|0
283261|5718290|0

After running, the ouput will create a file in the same directory cameras.csv.out with the following contents:

X|Y|Z|GapFraction
279890|5718602|0|32.4
279955|5718681|0|72.4
279880|5718759|0|12.45
279963|5718737|0|36.8
283261|5718290|0|12.4

Example 2: -loc cameras.csv cameras.csv with orientation , lens projection, setting output grid size to 1000x1000 pixels:

X|Y|Z|pitch|yaw|roll|proj|orast
279890|5718602|0|45.0|0.0|0.0|eqa|1000
279955|5718681|0|45.0|90.0|0.0|eqa|1000
279880|5718759|0|45.0|180.0|0.0|eqa|1000
279963|5718737|0|45.0|270.0|0.0|eqa|1000
283261|5718290|0|75.0|0.0|0.0|eqa|1000

After running, the ouput will create a file in the same directory cameras.csv.out with appended gap fraction values and ESRI ASCII grids with 1000x1000 pixels:

-orast: <pixel size of square grid>: default=180 exports reprojected shperical coordinates to a planar grid ESRI GRID ASCII format. Pixels represent the point counts. The cell values are the counts of points, scaled if one of -log or -db flags is provided. The name of the output files will be: Plot_<Number in 00X format><-log or -db if trasformation was used><-ort, -eqd -eqa -str depending on the projection chosen>.asc

-ori <0.0 180.0 0.0>- camera orientation, pitch, yaw and roll/tilt angles in degrees (optional). If not set, it implies an upward looking camera. E.g. 0.0 180.0 0.0 means a camera oriented towards the horizon looking south, not tilted. NB upward looking camera has pitch at 90 degrees corresponds to 0 degrees zenith angle - don't confuse the -zenCut value, that is in zenith angle.

-maxdist: <distance in meters>: default=1000.0 - any points falling outside this distance from the camera center will be ignored.

-log: converts pixel values, which represent point counts, to log10 scale (-orast must be also present) - formula is log10(pixelvalue+1). Cells with no pixels (value=0) are thus given log(1) and have value 0 also after transformation. This can be helpful as high zenith angles will obviously intersect a very high number of poitnts. Log-transformation can scale to better visualize results.

-proj: <str|eqa|eqr|ort|rct>**: default=eqa - different projections, respectively :

• str = stereographic
• eqa = equisolid / equal area (Default)
• eqr = equi-rectangular
• ort = orthometric

-db: converts to dB (decibel values) with -10*log10(pixelvalue/maxPixelValue). The pixel with most point counts will have value 1, the other will have positive values.

-weight <power value>: default=0.0 power of inverse distance weight. Each point that intersects a 1°x1° sector in the dome will be positioned at a certain distance that can be used to weight the value of the point. No weight=each point weights 1, i.e. if 10 points are in the sector, that sector will be occluded by 10 points. If weight=2 (-weight 2.0) is provided, each point will add a value of 1.0 * 1.0/pow(distance, 2.0) to the total. Default value is 0 because no weight is applied.

-zCam <height value in meters>: default=1.3m \n\t- height of camera - NB this is in absolute height with respect to the point cloud, so if your point cloud is normalized (e.g. a canopy height model) then 1.3m will be 1.3m from the ground.

-zenCut <Zenith angle in degrees>: default=89 \n\t- At zero (0) degrees zenith angle the direction points directly upwards, at 90 degrees it points at the horizong. Thus 90 degrees zenith angle potentially will intercept million of points: a smaller Zenith angle will ignore points lower than that angle (e.g. setting at 85 degrees will cut-off points that are below 5° from the horizon) .

1. a CSV file with the file name appended with .out and file contents of original camera locations file with appended column with gap fraction values.
2. If -orast is set, one ESRI GRID ASCII (.asc) text file that can be loaded in a gis software. One grid file per plot, named <plot. Center of grid is geolocated at camera position, but of course the size is not scaled.

Examples

lasPhotoCamSIM -i /archivio/LAS/las_normalized/forest.laz -loc cameras.csv -verbose

Will read all points from forest.laz and camera locations at cameras.csv file in current directory, providing verbose messages.

lasPhotoCamSIM -i /archivio/LAS/las_normalized/*.laz -loc camera.csv -log -weight 2.0 -verbose

Will read points in all LAZ files in folder /archivio/LAS/las_normalized/ and camera locations at cameras.csv file with verbose messages, and count points in falling in spherical sectors applying an inverse distance weight that is a power of 2. E.g. a point that is at distance X will count 1/X^2. A 5 cm minimum distance is applyed to avoid overflow of value if by chance a point is at 0.0m distance - this is reasonable as when camera is positioned, the operator will make sure that there not an obstruction right on the same position as the camera (e.g. right under a leaf).

lasPhotoCamSIM -i /archivio/LAS/las_normalized/*.laz -loc camera.csv -log -orast -weight 2.0 -verbose

Like above but will also create output grids in ESRI GRID ASCII format.

Using a normalized point cloud is more straight forward, if you want to focus on canopy light transmission.

This tool was created in the context of two projects,

CONAF - Postulación Fondo de Investigación del Bosque Nativo Title: Indicadores fenológicos y estructurales de alteración de hábitat en bosques de araucaria. Part of the activities was to assess estimations using real hemispherical photography, canopy-scope and photogrammetric point clouds. See paper....

VARCITIES - visionary solutions... In this context the light transmission in a garden is estimated to assess, with other variables, which areas are most suited for wellness of visitors.

Contact AUTHOR or visit https://www.cirgeo.unipd.it for more info and links to social-media, Facebook or Instagram.