Image Upload Tutorial - ngageoint/voxel-globe GitHub Wiki

Creating New Upload Session

Selecting “Upload” will take you to the upload session page, where you can start a new upload (Start New Upload).

images/Screenshot from 2016-10-27 19-17-53.png

To start a new upload, enter a name in the "New Session Name" box and choose the upload type, Image. There are type category types you must pick for for Image Uploads, the Metadata Type and the Payload Type.

Payload Types

The payload type refers to the image formats. Currently, two formats are supported.

  • Columbus Large Image Format (CLIF .raw files)
  • Image Sequence - Which can be thought of just a bunch of images, any images supported by tifffile, Pillow, and GDAL, (in that order)

Metadata Types

The Metadata is the additional information that you have describing the images. This can be anything from the header of images to an additional file, or additional files per image

All Metadata parsers initialize cameras at the corresponding coordinates pointing straight down. Orientation information is not used if supplied except in the VXL KRT case. No Metadata cameras default to GPS 0,0,0 and point straight down.

Since the association between metadata and payload data is not always straight forward, and since there may always be some additional metadata that is needed, a ingest_voxel_globe.json file can also be uploaded that contains additional information described here

For this tutorial, choose VXL KRT and Image Sequence, and name the session Providence

Click "Start New Upload" to start the upload session

Uploading files

From the file upload page, you can upload the files associated with the collection such as image and krt cameras

images/Screenshot from 2016-10-27 19-36-15.png

A test set of images can be downloaded at https://vsi-ri.com/data/downtown_providence.zip (271MB) and geo-register KRT cameras can be downloaded at https://vsi-ri.com/data/downtown_providence_cameras.zip

The “Select files…” button will pop up a file browser where you can select as many files at a time, to indicate that these are the files you want to upload.

Add additional files using “Select files…” again. “Upload Selected Files” starts sending all the selected files to the server. When all the files are complete, the “Process Uploaded Files” button is used to complete the upload session.

Let’s walk through this. Using Providence data, click “Select files…” First, select all the images and click “Open”.

images/Screenshot from 2016-10-27 20-00-19.png

Next click “Select files…” again, and this time select all the txt files and the ingest_voxel_globe.json file and click “Open”. If all the files are in the same directory, this can be done in one step instead of two. The upload page is flexible enough to allow this.

Now that all the files have been selected for upload, you will see all 363 files under “Files Selected for Upload" as seen here:

images/Screenshot from 2016-10-27 20-02-33.png

Next, click the “Upload Selected Files” button, to start the data upload. The previous image shows the list of files before upload has started. As files finish uploading, a green check box will appear next to their name:

images/Screenshot from 2016-10-27 20-10-00.png

If all the files have green checkboxes next to them, the “Process Uploaded Files” button will be enabled, and you can click it to start ingesting your files into the database. Click the button and return to the home page.

Clicking the Tasks link on the top right should reveal the upload status. giving you real time update as the upload continues:

images/Screenshot from 2016-10-27 20-12-06.png

Once the task has turned Green with Success, try the Image Viewing Tutorial to see the uploaded images.