Objects are tracked in real time in a composed video acquired by joining a plurality of videos. A grouping candidate determining unit extracts objects present within an overlapping area, in which pieces of frame data are overlapped, among objects that have been detected and tracked in each of a plurality of pieces of frame data that were captured at the same time as candidate objects. A grouping unit arranges a plurality of candidate objects of which a degree of overlapping is equal to or larger than a predetermined threshold as a group, and an integration unit assigns integration object IDs to groups and objects that have not been grouped.