BioFinder
VIDEO SEARCH
• Simultaneous recognition in multiple video recordings
• GPU technology, at speeds faster than real time
• Intensive, frame by frame analysis, without omissions
• Supports changes in facial expression and illumination, and the presence of glasses, beards, handkerchiefs or hats
• Capacity to enrol subjects through one or various pictures
• Alarms exportable to PDF format
APPLICATIONS
• Forensic analysis of video recordings
• Automatic classification of information from recordings
• Searches for suspects in large video databases
BioFinder is the solution from Herta Security that makes it possible to search for a specific subject, using facial recognition,
in one or various recorded videos. By working with GPU architecture, it manages to process at a high speed all the images
on the analyzed video.
BioFinder makes it possible to make analyses at speeds faster than real time. In this way it is possible to localize a
specific subject much faster than if the video is visualized at its original speed. BioFinder is ideal for forensic analysis or
classification of information and, amongst other applications, is used by the police to search for suspects.
• Simultaneous recognition in multiple video recordings
• GPU technology, at speeds faster than real time
• Intensive, frame by frame analysis, without omissions
• Supports changes in facial expression and illumination, and the presence of glasses, beards, handkerchiefs or hats
• Capacity to enrol subjects through one or various pictures
• Alarms exportable to PDF format
APPLICATIONS
• Forensic analysis of video recordings
• Automatic classification of information from recordings
• Searches for suspects in large video databases
| HIGH SPEED ANALYSIS OF VIDEO RECORDINGS | |
| Type of identification | Face, contactless, at a distance, on the move |
| Performance |
Unlimited recognition of faces on all the video frames. Analysis of up to 10 videos simultaneously |
| Database performance | One out of 200,000 user search in real-time with a conventional PC |
| Video capture resolution | UltraHD — 4K |
|
Resolution at detection |
Faces larger than 20 x 20 pixels |
|
Resolution at recognition |
Faces larger than 70 x 70 pixels, recommended 100 x 100 pixels |
| Resolution at enrollment |
Faces larger than 100 x 100 pixels, recommended 150 x 150 pixels |
| Face rotation | Optimal up to 30º, either horizontal or vertical |