Automatic software with neural learning of the scene, able to recognize and identify the presence of “human forms” in the scene framed by a camera. The recognition also works in case of limited visibility to some portions of the human form such as the head and part of the shoulders. The algorithm is particularly suitable for counting people in open places (squares, shopping centers, pedestrian streets, etc.), with a precision of 2-3%, depending on the size and depth of the scene.
