Python SDK for Bentley Systems Reality Analysis (RDA), Reality Modeling (CC) and Reality Conversion (RC) Services. It has been created and tested with Python 3.9.
This SDK is based on the APIs:
It also contains Reality Data (RD) utils to upload and download reality data. See Reality Data for more information.
-
You will need an iTwin platform account and a registered application to use this SDK. A general tutorial of how to register to the iTwin platform and how to create an Application can be found here.
-
If you don’t have a ProjectWise project you will need to create one. This video shows a step by step of how to do this at CONNECT Center.
When creating an Application you will need to include scopes so that the Application has authorization to start jobs, etc. To be able to use the Reality Data Analysis and ContextCapture services and also have the possibility to upload and download data you should choose Reality Capture, Digital Twin Management and Reporting & Insights as API associations. Your application should have at least contextcapture:modify, contextcapture:read,realitydataanalysis:modify,realitydataanalysis:read, realityconversion:read, realityconversion:modify, realitydata:modify and realitydata:read as Allowed scopes for this SDK to function properly. Other scopes won’t interfere with the service.
By default, the token factories provided expect as redirect URI http://localhost:8080/sign-oidc.
You may choose any Application type but be conscious the authorization process is different between them. An explanation of each authorization code flow process can be found here. This SDK includes an example of a token factory for each type of Application. You may code your own factory for authorization tokens but you will have to implement the functions present at AbstractTokenFactory so that your code integrates properly with this SDK.
After creating an application and a ProjectWise project if you didn't have one, you can either:
- Clone this repository and add its path to your PYTHONPATH environment variable;
- Download the last released version wheel and install it using
py -m pip install **path_to_the_wheel**; - Install it directly from the last github version by using
py -m pip install git+https://github.com/iTwin/reality-capture.git#subdirectory=python
and you are good to go.
Necessary Python modules can also be found at the requirements.txt file.
requests
bottle >= 0.12
oauthlib >= 3.2
requests-oauthlib >= 1.3.1
azure-storage-blob >= 12.9
In order to get consistent context scenes and orientations on Context Share, you have to upload the data referenced in those files first. Indeed, context scenes and orientations contain local paths to image collections, point clouds, etc. However, when the context scene is uploaded, these paths are not valid anymore. To have a working context scene or ccorientation when uploading from your local machine, you have to:
Upload references first. When uploading an image folder, images should be at the root of this folder, and at the same level (no sub folders). The same thing applies to orientations, files should be uploaded before you upload the ccorientation file.
Change the references. You can do that by hand by saving the ids of the files you uploaded to the cloud, opening your context scene or ccorientation file in a text editor and changing the paths to those ids. You can also use the ReferenceTable object to save paths and ids while uploading reality data and the upload_context_scene or upload_ccorientation functions to replace paths automatically before uploading those types of files.
To run an example file you will need to change the config.py file: client_id should be the id of your application and project_id the id of your ProjectWise project. To retrieve the id of your project, click on the ContextShare button bellow the name of your project at CONNECT Center and copy the contextId string on the address bar. Don't forget to also change the secret variable if your application has one.
You will also need to change the variables at the beginning of the example file you’re running so that they reflect the files you want to use and where you want to save the results. You can download the detectors available as well as example datasets here. Don't forget to first open context scene files with a text editor and replace the reference paths to the path of the images in your machine so that they point to the local path where you saved the images.
https://developer.bentley.com/apis/realitydataanalysis/overview/
https://developer.bentley.com/apis/contextcapture/overview/
https://developer.bentley.com/apis/realityconversion/overview/