Hi Agisoft Team,
It would be very useful if there was a way to increase sensitivity of coded target detection and ask the user to confirm if each questionable detection is a target. You could show the user the predicted coded target shape alongside the questionable target detection in the photo and the user could confirm if it is a target. Through this data you could also train a neural net to improve automatic detection. You could have an option for the user to contribute the training data to you to refine the model further. This training data could be a small crop of the target in the original image with appropriate tags.
Cheers.
Topic: User Assisted Coded Target Identification (Read 1651 times)
