Abstract: Provided is an intelligent image analysis system metaverse system for analyzing an object by intelligently processing camera images of a plurality of spatial zones in each of which at least one camera is installed. The intelligent image analysis system includes body shape encoders configured to detect an object in a camera image of each spatial zone and encode the object into a shape identifier (ID) on the basis of features of a body shape, a spatial-zone-specific identification manager configured to manage the shape ID together with camera information and time information, an identical shape ID extractor configured to extract identical shape IDs from different spatial zones on the basis of a list of shape IDs which are encoded according to each spatial zone, and an identical object determiner configured to determine a shape ID related to the same object using a distance between two spatial zones and imaging time information.

Abstract: Provided is a metaverse system. The metaverse system for providing metaverse content virtualizes a target space as a digital twin but statistically applies attributes and action patterns of objects in the target space to virtual objects in a virtual space, thereby reproducing the environment, atmosphere, and the like of the target space without accurately copying the attributes and action patterns of objects.

Abstract: A video analysis system is a system in which a video is analyzed through an edge device installed in each site to generate a security event, and when the edge device continuously generates erroneous security events, the video is analyzed through another edge device or a cloud server in parallel, and then a deep learning model enabling the edge device not to generate an erroneous security event is found to change the edge device, thereby maintaining high video analysis quality.