Knowledge Base
All you need to know about how to use The Biodrone Portal to count and categorize objects, calculate values, detect damages and identify risks to make insight-driven decisions.
Title
Lorem ipsum dolor sit amet, consectetur adipiscing elit. Quisque rhoncus erat quis metus cursus, quis convallis orci rhoncus. Nulla facilisi. Pellentesque volutpat congue libero varius euismod.
Video Tutorials
How to count trees with AI
Use the new filter function
Frequently Asked Questions
You can upload any data from optical sensors. For creating orthomosaics, upload RGB data with sufficient overlap. For the single image feature, you can upload any data from optical sensors as long as the image format is JPG.
Processing can take several hours and depends on the number of images, resolution, and area size. Poor overlap, blurry, or uniform images may increase processing time. To speed up the process, ensure high-quality images with good overlap and resolution.
Re-capture your data with higher overlap, and make sure the images are clear, properly exposed, and not blurry.
Absolutely! BioDrone can be used for a wide range of applications, such as detecting weeds, mapping invasive species, counting animals, and inspecting infrastructure, as long as the objects of interest are visible in your images.
Choose based on your mission: Use Object Detection if you want to detect and count objects. Use Segmentation if you want to classify areas and measure the extent of specific types.
Yes, you can run multiple models simultaneously on the same project, but you cannot run the same model at the same time.
Improve accuracy by using the Train AI function on multiple datasets or using the Improve Accuracy function on previously processed data.
If you make an error while annotating, you can correct it by removing or moving the annotation before clicking Continue. However, once saved, annotations cannot be modified.
Yes, project and image limits depend on your subscription type.
You can share data by downloading analysis results and orthomosaics, or use the WMS function if your subscription allows it. Export data using the export function button to download the files.
Yes, you can train your own models by uploading and marking the objects you want the model to learn. Once training is complete, you can use the model across all your projects.
Check that the training data is similar to the data you’re analyzing (e.g., don’t train on green apples and try to detect red apples). Ensure enough training data is provided; consider adding more annotations and retraining the model.
Yes, you can run the analysis on a small section of your map first to review results. When you run the model on the full dataset later, the results will merge, or you can delete the initial detection.
Yes, BioDrone offers support for model training, detection, labeling, and data collection. Contact us at post@biodrone.ai.
Yes, if you are part of an organization, you and your team can work on the same models and collaborate. However, you cannot train on the same model simultaneously.
BioDrone uses AWS as its cloud service provider, incorporating the latest in data security measures offered by AWS.
You need a web browser and a good internet connection for uploading data. There are no specific hardware requirements.
BioDrone currently requires an internet connection; there is no offline version available.
You can change your subscription on your profile page within the platform.
BioDrone exports analysis results as GeoJSON and KML files, and orthomosaics as GeoTIFF files.