Robust Moving Object Detection at Distance in the Visible Spectrum and Beyond Using A Moving Camera

Abstract

Automatic detection of moving objects at distance and in all weather conditions is a critical task in many visionbased safety applications such as video surveillance and vehicle forewarn collision warning. In such applications, prior knowledge about the object class (vehicle, pedestrian, tree, etc.) and imaging conditions (shadow, depth) is unavailable. What makes the task even more challenging is when the camera is non-stationary, e.g., mounted on a moving vehicle. The essential problem in this case lies in distinguishing between camera-induced motion and independent motion. This paper proposes a robust algorithm for automatic moving object detection at distance. The camera is mounted on a moving vehicle and operates in both day and night time. Through the utilization of the focus of expansion (FOE) and its associated residual map, the proposed method is able to detect and separate independently moving objects (IMOs) from the "moving" background caused by the camera motion. Experimentations on numerous realworld driving videos have shown the effectiveness of the proposed technique. Moving objects such as pedestrians and vehicles up to 40 meters away from the camera have been reliably detected at 10 frames per second on a 1.8GHz PC.

Motion object detection by three-view constraint using moving camera

“…Through the above experimental data, the three-view distance constraint which this paper presents can supplement the epipolar constraint and change the surface degeneration to the line degeneration to detect the motion object by a moving camera. The residual image can be obtained by the residuals of the feature points d p using the method [26] proposed. The residuals of the feature points d p are shown in Fig.…”

“…In order to perform the further tasks like object tracking, we need to segment pixels of the moving objects in the image. [26] proposed a method where it used the residual of the feature points to obtain the residual map for all pixels in the image. For a M Â M (M ¼ 7 in this paper) window in the image, the residual of each pixel in this region is set to the average residual of the interest points in this region.…”

Detecting Flying Objects Using a Single Moving Camera

“…Some methods combine both optical flow and background subtraction algorithms [27], [28]. However, in our case there may be motion Figure 3: Object detection pipeline with st-cubes and motion compensation.…”

Video saliency detection incorporating temporal information in compressed domain

“…Picking out relevant parts of a scene as visual attention regions is a hot research topic and has a wide application prospect, such as content-based image and video retrieval [1], perceptual video compression and coding [2], object detection and segmentation [3][4][5], video analysis [6]. Recently, to tackle these information overload problems, numerous computational saliency models are proposed to simulate biological vision systems by researchers in physiology, psychology, neural system, and computer vision.…”