A quick start guide for those new to EVA.
<aside> 📘
Recommended flow: Proceed in the order of permission/license/RTSP preparation → camera registration → single-purpose scenario creation → object/area setting → monitoring → notification/feedback verification.
</aside>
EVA not only provides the function of finding an object in camera images, but is also an operational tool that determines based on the scenario whether the object found actually requires response. So, when setting up for the first time, it is important to check that video, object detection, situation determination, and notification are sequentially performed, rather than assuming that the camera is connected simply because it is connected.
Start with one representative camera and the clearest scenario. Below, we first look at the permissions and preparation conditions required for operation, then understand the process by which EVA creates notifications, and then follow the actual setup and verification procedures. Once you experience this flow, expanding to other cameras and scenarios will make it much easier to find the cause if something goes wrong.
Changing settings may require Admin or Manager privileges.
EVA receives camera images and first checks the object's location and reliability with a vision model, and then the agent reviews the situation and exception conditions of the scenario to determine whether or not it is actually notified. Therefore, seeing an object on the screen and firing a notification are not the same steps.
When first operating, it is safer to start with one camera and one scenario and expand the scope after checking that object boxes, agent judgments, and notification cards appear in the expected order.