geomapi 0.0.14

A standard library to manage geomatic data

GeomAPI

A joint API to standardize geomatic data storage and processing.

[[TOC]]

Installation

Use the package manager pip to install geomapi.

pip install geomapi

Documentation

You can read the full API reference here: Documentation

Usage

The main use of this API is importing standardized RDF data into easy to use python classes. These python classes have a number of fuctions to analyse, edit and combine the most common types of data including:

• Images (pinhole or panoramic)
• Meshes
• Point clouds
• BIM models

Variable Standards

This library relies on a RDF based standardisation of data storage, Below are some of the more important variables:

• cartesianTransform: A 4x4 transformation matrix
• Paths
  • path: The absolute path of the Asset.
  • graphPath: the absolute path of the graph folder

Development

Testing the package is done in the tests folder with:

from context import geomapi

Work method

• nieuwe code comitten, aanpassingen brachen of aan MB vragen
• nieuwe dependencies worden gestemd

input protection/ error handling

• try catch blokken
• zoveel mogelijk protection binnen functie
• type hinting input + output

documentatie (uitleg, eenheid , datatype)

• uitleg bij functie
• uitleg bij parameters
• type hinting input + output

standardisatie

• beperk custom classes waar mogelijk
• beperk aantal argumenten
• gebruik zoveel mogelijk standaardconcepten
• name is zonder speciale tekens of spaties (door windows en RDFlIB)

PYTHON

• module/files: allemaal kleine letters
• classes: Capital elk woord
• functie: allemaal kleine letters met underscores
• variables: kleine letters, allemaal aan elkaar, 2de woord capital
• niet teveel afkortingen

FUNCTIONS

• histogram in a certain direction
• split horizontal from vertical points
• create virtual image
• triplestore
• documentation
• fix graphPath constructor
• crop point clouds
• collision point clouds
• make imagenodes from vlx metadata

KWALITEITSANALYSE

• rendering van KWALITEITSANALYSE

CROP MESHES

• check mesh collision functions

• crop_geometry first run a check if the Bboxes intersect => saves alot of intersection calculations

• crop_geometry add non_overlapping bool parameter that culls mesh with each selection => should speed up process

• crop geometry functions have a lot of deepcopies which hurt performance

• intersection based on convex hull

• crop geometry 1m30 on 600 nodes with 40k mesh => for 2M mesh and 6000 boxes this would take 1h30

• trimesh mesh intersection 2m17 for 600 nodes on 40k mesh =>for 2M mesh and 6000 boxes 1000min

CROP PCDS

• Python has to memory allocation protection :'(

• create CC script thats loads pcd's and outputs pcd

• segment large pcd's into regions

• BBoxes pcd are to large

• crop BB 4m30s for 6x10M pcd 600 boxes => for 45 clouds and 6000 boxes (normal project) this would take 5h30

• segmentation based on Eucl. distance 5ms for 6x10M pcd 600 meshes => for 45 clouds and 6000 boxes (normal project) this would take 6h

GEOMETRY

• protect oriented bounding box en bounding box tegen puntenwolken met te weinig punten (coplanair of 1pnt)
• e57 scanheader also contains oriented bounding box?

LINKED DATA

• functie node.get_oriented_bounding_box()

IMAGENODE

• implementeer xml import

ENVIRONMENT

• python-fcl is causing crashes!

Licensing

The code in this project is licensed under GNU license.