could you please clarfiy your use case and provide more details regarding the 'preprocessing of the camera flow and feed'?
Thx and greetings
At the moment, we don't offer a way to preprocess the camera stream in this way. However, you can achieve something similar with an Input Plugin, for which we have samples and documentation for both iOS and Android.
The idea behind using an Input Plugin is that the device camera can be controlled from outside of the SDK. This way, you can get the camera frames directly from the camera, preprocess them as needed and then forward them to the SDK to run image tracking.
Please let me know if you have any additional questions.
Also, I want to ask about the number of the photo inside the marker sets. Is that the more the better? While we put more pictures into the marker, the flashing of the video appeared.
Are you still referring to image tracking, or is this question related to object tracking?
The question is related to image tracking, thank you!!
For image tracking, there should be a single image in the target collection for every you plan to recognize and track. There shouldn't be duplicates, or different versions of the same image. That can actually have a detrimental effect to the quality of the tracking.