If a swimming pool is a predictable and limited environment, then open water is much more complicated. Lakes, rivers, and the sea have variable conditions: wind, waves, sun glare, and changing transparency. The camera and algorithm have to deal not only with the image of a person against the background of water, but also with a multitude of dynamic factors that can mask or distort the signal.
In addition, in open water, people can be very far from the camera, partially hidden by waves, changing their poses and behavior. What is easily captured from above in a pool becomes a highly complex task in natural conditions.
Nevertheless, technology is advancing. Systems are being developed that take wave dynamics into account, analyze movement in different frequency ranges, filter out glare, and use multiple sensors at once, such as thermal imagers. This makes it possible not only to “see,” but also to truly recognize risk, even in complex and changing environments.