Fusion of Thermal and Visible Imagery for Effective Detection and Tracking of Salient Objects in Videos

Abstract

In this paper, we present an efficient approach to detect and track salient objects from videos. In general, colored visible image in red-green-blue (RGB) has better distinguishability in human visual perception, yet it suffers from the effect of illumination noise and shadows. On the contrary, thermal image is less sensitive to these noise effects though its distinguishability varies according to environmental settings. To this end, fusion of these two modalities provides an effective solution to tackle this problem. First, a background model is extracted followed by background-subtraction for foreground detection in visible images. Meanwhile, adaptively thresholding is applied for foreground detection in thermal domain as human objects tend to be of higher temperature thus brighter than the background. To deal with cases of occlusion, prediction based forward tracking and backward tracking are employed to identify separate objects even the foreground detection fails. The proposed method is evaluated on OTCBVS, a publicly available color-thermal benchmark dataset. Promising results have shown that the proposed fusion based approach can successfully detect and track multiple human objects.

Enhancing Deep-Learning Object Detection Performance Based on Fusion of Infrared and Visible Images in Advanced Driver Assistance Systems

“…VIS imagery in RGB has good distinguishability in human visual perception, but it is susceptible to shadows and lighting noise. In contrast, thermal IR images are less sensitive to these noises and can provide additional information for warm objects at night [8]. Therefore, fusing these two kinds of complementary images may provide a useful solution to the object detection problem.…”

Cognitive Fusion of Thermal and Visible Imagery for Effective Detection and Tracking of Pedestrians in Videos

“…In the past decades, detection and tracking of video objects has always been a major task in the computer vision field [1][2][3]. As one subset of video object tracking, pedestrian detection and tracking has drawn massive research attention and been applied to many applications such as visual surveillance [4][5][6][7][8], driverassistance systems [9][10][11], human activity recognition [12][13][14], and others [15,16].…”