If you have uploaded images on an S3 bucket, you can import these images to Arkindex in a few steps.
In order to import images from a bucket into an Arkindex instance, you first need to ask an administrator of that instance to grant you access to the S3 import feature.
Your images must be imported into a project: you can create a new one by clicking the New project button in the top-right corner of the Arkindex homepage. To create a project, you only need to fill in a project name, and optionally a description.
You can also import your images into an existing project.
In order to access the S3 import interface, you need to navigate into your project, and from there click on Import files from S3 in the Actions dropdown menu.
A form asks you to fill out 3 fields:
First, you need to select the bucket your images are in from a dropdown.
If your bucket isn't listed, you need to ask an administrator of the Arkindex instance to give Arkindex access to that bucket.
You can also specify a key prefix to only import some of the files in the bucket.
For example, if you enter the prefix folder1/myfiles
then files with names such as folder1/myfiles/1.jpg
will be imported, as would a file named folder1/myfiles.png
.
Finally, you can select an existing folder in the project for the images to be imported in.
If you do not select a folder, then a new one, containing your images, will be created.
By opening the Advanced settings dropdown, you can specify the types of the elements that will be created by the import. By default, Page elements are created from the imported images, and the created folder is of type Folder.
Once you have filled out this form, you can click the Import button to start the import.
Clicking the Import button takes you to the Process status page, from where you can monitor your import's progress.
You can leave this page and find it again from the Process list, which you can access by clicking Process in the main menu bar at the top of the website.
Once the import is complete, you can browse the created elements, annotate them, or process them using machine learning workers.