I printed out the markers on A4 and gave it a try. Since A4 isn’t that sturdy the marker got distorted and both examples had a lot of trouble picking it up. Also I tried it at night in the study with a single light on. This resulted in a glare on the marker and of course this interfered with the recognition. The low light conditions didn’t help much either, it seems a low(er) contrast really throws the marker recognition off. All in all AR software seems to be quite sensitive to imperfections.<\/p>\n
Of course this could stop me from trying something myself and see why exactly it is so sensitive. [Click the image above to run the example, I’ve used the same marker as the Boffswana example.]<\/em><\/p>\n If you run my test, you’ll see that it’s quite sluggish as well, you’re lucky if you get 6+ frames per second. This is due to the marker detection method ( FLARSingleMarkerDetector.detectMarkerLite() ). Sometimes this method takes up to 1300 ms to detect the marker, on average it takes about 150 ms.<\/p>\n The Flash community has always been really fast on picking new technolgy up, but I’m afraid this is a little bit too high end for Flash at the moment.<\/p>\n Some possible solutions:<\/strong><\/p>\n This seems to be the new buzz word on the web. After seeing a couple of examples from Boffswana and Toyota I wasn’t that convinced. The response usually isn’t too good and the quality seems to lack a lot. I … Continue reading
\nSo what I did is combine the FLARToolkit<\/a> with Flint particles<\/a>.<\/p>\n![\"Augmented](\"http:\/\/ansuz.nl\/toys\/ar\/ar_test_01.jpg\")
<\/a><\/p>\n\n