TraceSearch

TraceSearch allows you to draw a box around a specific person or vehicle and find them as the cross your other cameras. This makes finding someone’s specific path throughout your facility simple and easy! What used to take hours can now be done in minutes with TraceSearch!

Thanks to next generation processing on our NVRs - you can find a persons path through your facility from entrance to exit. Answer questions like “Which door did they go into”, “When did they arrive and leave” easier than ever.

TraceSearch requires an NVR with a next generation processor. Currently, only the Admiral v4 series supports TraceSearch. You also need to have cameras that have built in AI person/vehicle detection and have the analytics setup in order for TraceSearch to work.

Q: Does the TraceSearch cost?

A: No - they are built into the NVRs and have no recurring/monthly fees to use them.

Q: How does it work?

A: TraceSearch uses color and machine learning to find similar looking people and vehicles. This largely utilizes clothing colors, shapes, etc.. It does not utilize facial recognition.

Q: What if someone changes their clothes?

A: A: If someone changes their shirt to a different color - most likely the similarity score will be dramatically lowered or even not match at all, meaning you’d need to do a search for both color shirts.

Why do I need a v4 NVR?

A: V4 NVRs feature a brand new, powerful AI chipset that is responsible for the TraceSearch process. Previous NVRs do not have the chipset.

Q: Does this work with third party/ONVIF cameras/Older SCW Cameras?

A: No - TraceSearch requires SCW cameras that support built built in AI person/vehicle detection. Third party cameras or SCW cameras that do not support the built in AI will not show up on TraceSearch results.

Q: How do I get the best results possible.

A: Make sure you choose a still of the person/vehicle that is clear and sharp - since the NVR is searching based on that image, the better the initial image the better the results. Low quality images, for example someone very far from the camera will likely produce poor results. You can also adjust the similarity bar to adjust the results.