Trying to understand the head mounted camera feature. I wanted to mount a small camera on the laser head with hopes to improve the print and cut workflow and feature. Unfortunately I can’t find any good information on the topic and even any recommended cameras for head mounting.
We need a feature to print 5mm or so solid circles and the head camera to detect those automatically after the first one is manually set. This type of feature has been out there for many years.
Also, what is the point of dual cameras I read about. Not much info? Can someone side of Lightburn expand on the subject. People are out here paying $1,500 for third party software just to improve registration on print and cut. We need to advance this feature asap. Behind the times for sure. Please any help on the camera subject is appreciated.
Print and cut does not need a camera at all. That’s the first thing. I don’t think the camera feature was designed to be used for it.
The head-mounted camera is just using multiple shots to create the full overlay image of the work area. No features are captured related to the laser head.
Dual cameras enable users to use two cameras next to each other to cover the full workspace when a single camera can’t catch everything, as far as I remember (I might be wrong).
So, you are actually searching for a way to take an image of objects and place the engraving on it automatically?
No. I’m looking for a common method to make print and cut more automatic. Other software program accomplish this with a head mounted camera. It’s used to register mark 1 and then it automatically registers the remaining marks which improves workflow and accuracy.
It’s not there yet. We need a on head camera that can read rego marks and automatically align printed image. Apparently Lightburn were looking into this but not sure how high it is in their priority list. I am patiently waiting for this function as we now have a Titan with on head camera and Thunderlaser’s software is….well…..basic at best.
Image recognition is surprisingly difficult and a true DNN (dynamic neural net) requires a powerful GPU, which not everyone has.
As far as I know, many consumer brands that offer a similar feature use a cloud-based service, which we won’t want to do. Some manufacturers also state that the feature can only detect black objects.
A reliable object detection system will require high-quality (industrial) cameras and good consistent lighting with a contrast to the background.
We’ll definitely keep an eye on the possibilities. With the new camera system in version 2.1, we’re well-positioned to implement such advanced features in the future.
This would be great. There are a lot of brands that use the same sort of system and run quite well on basic computer hardware. Iecho, Multicam, summa and Graphtec too bname a few. We are not asking for the software to look at the images on the printed item but 2 or more registration marks (ideally at least 4) that can be used to register the cut to the print.