Germany develops face recognition technology in the dark

German scientists have recently reported a successful development of a new face detection technology even in the dark.

Facial technology in the dark

New identification technology uses a special software based on the heat radiating from the face as a basis for recognition. Most traditional face recognition software today uses two-image comparisons in bright enough environments, so it is unlikely to be effective in low light conditions.

Computer scientists Saquib Sarfraz and Rainer Stiefelhagen from the Karlsruhe Institute of Technology in Germany have created an intelligent analysis system capable of analyzing dozens of infrared images of the human face and then comparing them with Dozens of other shots were taken with daylight.

The comparison was made with a computer program that was based on a highly specialized neural network that could be designed and imitated in the same way as a human brain function.

Sarfraz and Stiefelhagen explained the operation of this network by analyzing 4,585 photographs taken in both infrared light and variable conditions. The system then produced a pretty quick result in just 35 ppm.

Today, though, the technology still needs to improve on how it works when images depicting human face temperature can fluctuate depending on weather conditions. That is not to mention many cases after the exercise will also show different temperatures compared to normal.

But talking about the prospects of this technology, scientists are quite convinced that the accuracy rate is more than 80% when comparing a photo describing face temperature with more pictures. Good lighting conditions. But this ratio can be reduced to 55% by comparing a variable image for each object.

The technology promises to be an effective means of helping security experts and law enforcement agencies investigate, disrupt and secure their security.