This article explains how to process files with Video Recognition Gateway (VIRGO) video feeds. While it's possible to process files in the Video Feeds window in the SAFR Desktop Client, it can be useful to process files in the background using VIRGO (available thru the Video Feeds window) for the following reasons:
- Processing a file in a loop for demo purposes
- Processing many files in parallel (future articles will demonstrate how to load many feeds quickly)
This article will explain how to create a video feed from the Camera Analyzer in the SAFR Desktop Client and then push those settings to the Video Feeds window. Following that it will explain how to change the input from camera to file.
1. Create Video Feed by following this process: Creating and updating feeds from Camera Analyzer
2. Changing Feed from Camera to File Source
Open Video Feeds Window and expand the feed you have added under the "feeds" list as shown below:
Edit the Feed to change feed to a file type as follows:
a. Mouse over the feed name and Click the "+" to add an attribute.
If needed pre-type to filter for the attribute in question. In this case all the attributes you need are right there at the top under "input.".
After editing, you should see the following settings among others:
- input.video-clock.enabled impacts the playback speed. If "True", playback is in real time (good for demos or simulating for testing). If "False", playback is as fast as SAFR can process (good for processing historical videos for forensic analysis).
- input.loop causes the file to loop. If set to "False", video will stop at end and you will have to disable / enable the feed using the "enabled" property to restart it.
- input.type=file tells SAFR to look for a file input
- input.stream.url should be set to the path to the file. Path is platform specific. Use "\" for Windows and "/" for Linux. DON'T forget to add the file extension. (Sometimes it's not visible in Windows so can be easy to miss.)