Yes, as JPEG format

The higher the image resolution - the better the detection will work. In general images with GSD of 2-5 cm/pixel should be sufficient. To detect bigger objects like cars, solar panels, vegetation areas, etc it can be sufficient with 10-15 cm/pixel GSD.

Yes. In our training materials, you can find sample datasets here: https://learn.atlas.ugcs.com/docs

Yes, on paid Cloud Plan and On-Premise it is possible to export detectors as files. To run this detector locally you need ATLAS On-Premise

We do not ship detectors out of the box, because for many tasks quality depends on many conditions. But to train your first detector you may need 30 minutes. It is really quick.

When detection is finished we cut results below 80% of confidence. In general, a lot of things depend on the annotation quality. Please follow our guidelines https://learn.atlas.ugcs.com/docs

Images with detected objects can be downloaded as a ZIP archive or CSV list
From it is possible to download GeoJSON data for external GIS systems and PDF report with area screenshot and identified objects counts and area sizes

ATLAS supports multiple users in one organization. You can invite them and open access to a specific project. It is also possible to share the project as a public link, not requiring authentication.

Yes. Data can be captured with any tool and then processed into orthomosaics

We provide powerful object detection AI but at the same time allow to structure data into projects and track changes on a timeline

Yes. ATLAS has also a photogrammetry engine under the hood

Yes. We support GeoTiff format

Depending on the plan. We do not charge for training. We charge for detection. You can divide plan monthly capacity by image size to see how many images can be processed

Yes. The video should be in Full Motion Video (MPEG-TS) with GPS coordinates. One can record such a video with the UgCS Enterprise video recording package

Yes. On-Premises option allows applying branding for a local installation

Yes. Please contact our support@ugcs.com for detailed information

Yes. Now it is possible to export GeoJSON, Shortly we will support DXF export

Size in megapixels can be calculated as the multiplication of image width and image height in pixels. To calculate how many pixels you will require use the ATLAS calculator >>>

GB stands for Gigabytes. The amount of data that you can store in our Cloud according to a payment plan.

Please check our Pricing page with the comparison of Plans

Yes, ATLAS can be installed on your own hardware. Please contact ilze@ugcs.com for On-Premises options.

Currently ATLAS has English, Spanish and Chinese localisation

Yes! We offer 14 days of a free trial. To subscribe please see the Pricing section

The subscription plans can be paid with

• payment card (Visa, Maestro, Master and American Express),
• PayPal
• AliPay.

Yes! No hidden cancelation fees - cancel your subscription anytime and use the resources which are already paid till the end of the period.

Yes. This is the ATLAS On-Premises installation. Please contact ugcs@ugcs.com for pricing options and technical details.

The configuration depends on the amount of data you want to store and process, but the minimal configuration may consist of 3 nodes:

• Main server: Linux Ubuntu, Core i9 (6 cores), 32Gb Ram, 512Gb SSD, 1Gbit Ethernet
• Server with GPU: Linux Ubuntu, Core i9 (6 cores), 64Gb Ram, 512Gb SSD, 1Gbit Ethernet, Nvidia RTX2080Ti
• NAS: depends on the storage size

Yes. ATLAS supports Amazon AWS. For deployment, you'll need access to EC2 instances and S3 storage

Partially. Internet connectivity is required to activate the license and access base maps from OpenStreetMaps. But base maps can be substituted with any available tile server. We never send your data from ATLAS On-Premise to our servers.

No. The detectors are not shared between organization.

Data of each organization is isolated and available only for members of your account. Access to members to your account can be provided either by creating login and password or with Google authentication.