Noise Reduction Test

If we are able to create an app that combines Suan Young's background error correction with software noise reduction, I'm sure we would be able to get some very impressive photos out of the Therm-App.

 

Right now, Therm-App images are noisier than even the lower end FLIR cameras (ie. E4/E8). I don't think the microbolometer inside the Therm-App (ULIS Pico384P) is inherently so much worse (if worse at all) than FLIR's. I think it's more a case of a combination of a lack of calibration shutter (hence no background error correction) and very little image processing implemented by Opgal.

 

If we look at the 2nd gen FLIR One, they recommend at least a Galaxy S5 / Nexus 6 (2.2Ghz+ Krait Quad Cores) or good video recording (source: www.flir.com.hk/flirone/content/?id=69420#compat), where even a Nexus 7 (1.7Ghz Krait Quad) supposedly produces slow video recording. I can only assume from those system requirements that the FLIR One does very intense image processing, which the Therm-App definitely lacks.

Done