These two drivers (imgcmp, imgsave) typically used together do following things:
1. imgcmp compares image frames in order to detect changes (for example, when someone moved in front of the camera), changes are denoted by differing timestamps and publishing logical state through the dio interface. Diagnostic image can me obtained from optionally provided camera interface.
2. imgsave stores camera image as files (xyRGB or JPEG) in directory tree organized by camera name, date, time
3. imgsave can print additional informations on saved images (camera name, date, time when image was stored) - this will ease working with surveillance materials.
4. imgsave stores images only when timestamp differs: using this driver with imgcmp, diskspace can be saved as images would be stored only when something happeds.
These two drivers are part of some bigger project that I'm working on in my spare time. More interesting thing may leak from it here in some near future.