Comparison of thermal cameras for protection by video analytics

On last March 29thEVITECH was organizing a night comparison session between several thermal camera models from the market.

The purpose of the test was to demonstrate and compare the following products in the same situations and conditions, outdoor in the campain, at the same time. Evaluated products were : the new camera DRS WatchMaster IP Elite 40 mm (sensor 320 x 240, FOV 9°), the latest  AXIS Q1922-E 19 mm camera (sensor 640 x 480, FOV 32°), presented to the market on March 28 (!), the FLIR F-313 camera (sensor 320 x 240 pixels, FOV 13°), the FLIR PT307 tower (one thermal sensor of 320 x 240 pixels, FOV 7°, one color sensor), and last the small  SAGEM Sword camera (sensor 352x288FOV 45°).

First, a presentation was done at EVITECH's resort in the south of Paris, with the products presented on the table indoor, with all the images in parallel on the same screen.

The constrasts were particularly observed on the Sagem camera, which has no settings at all. 

The DRS Elite camera is of simple use, with an optics focus that can be set by hand, and an embedded configuration website (IP interface).

FLIR sensors have a fixed and pre-set FOV and focus. Settings are also made by IP. The 7° FOV provides to this camera the best performance in distance, among the different cameras evaluated.

The image of Q1922 AXIS camera  was the highest resolution one, with its 4CIF sensor, which also provides a well contrasted and low noise image.

Then the participants had a quick meal before leaving to the evaluation site, outdoor in the fields.

The weather was dry and fine, temperature from 15°C to 10°C, and hygrometry 60%.

The participants could admire on their arrival the "thermo-quad" from France Infra Rouge. They could compare the products in forest and fields situations at different distances : image quality, optical noise, NUCeffects, resistance to various hiding techniques (covers, thermal protection, paperboard, ...), all was tested until 22 H. Jaguar detection was set with similar "low" settings on every cameras so that detection could illustrate the image quality, by detection or non-detection depending on image quality.

Participants mentioned their satisfaction and encouraged us to reproduce this type of event. Please notify us if you would be interested to participate to a future event of this kind.