Detecting moving objects from omnidirectional dynamic images based on adaptive background subtraction

This paper describes a new method for detecting moving objects with an omnidirectional camera. The proposed method uses a fixed omnidirectional camera which can acquire images surrounding the camera and a background subtraction technique based on adaptively updating background images. In subtracting the background image from the input image, we consider the oscillation of intensity so that the proposed method can robustly detect moving objects under the condition that there are fluorescent lights and incoming lights from windows. In experiments, moving objects have been successfully detected in indoor environments. 2.1. Outline of the method The proposed method uses an omnidirectional camera and a background subtraction technique. The fixed omnidirectional camera acquires omnidirectional images surrounding the camera. First, we estimate the background image and amplitude image from the first several frames of the input omnidirectional image sequence. After the background image and amplitude image are estimated, we detect moving objects from the input images, the background image and the amplitude image, updating the background image and the amplitude image. Next, we generate common perspective images of moving objects, calculating fan-shaped bounding-boxes which enclose the detected moving objects in the omnidirectional image