Sun orientation estimation from a single image using short-cuts in DCNN. (February 2019)

• Record Type: Journal Article
• Authors: Jin, Xin; Sun, Xing; Zhang, Xiaokun; Sun, Hongbo; Xu, Ri; Zhou, Xinghui; Li, Xiaodong; Liu, Ruijun
• Is Part Of: Optics & laser technology. Volume 110(2019)
• Page Start: 191
• Page End: 195
• Publication Date: 2019-02
• Subjects: Sun orientation estimation -- Inserting 3D models -- Single outdoor image -- Augmented reality; Optics -- Periodicals; Lasers -- Periodicals; Electronic journals; 621.366
• DOI: 10.1016/j.optlastec.2018.08.009
• Languages: English
• ISSNs: 0030-3992

Abstract

Highlights: We propose an end to end way of outdoor sun orientation estimation via a novel deep convolutional neural network. Our method needs only a single image. Our method can directly output the orientation of the sun from an outdoor image. The estimation accuracy of our method outperforms the state-of-the-art DCNN based method. The method has good application potential. Abstract: The illumination effect is essential for the realistic results in images which are created by inserting virtual objects into real scene. For outdoor scenes, automatic estimation of sun orientation condition from a single outdoor image is fundamental for inserting 3D models to a single image. Traditional methods for outdoor sun orientation estimation often use handcraft illumination features or cues. These cues heavily rely on the experiences of human and pre-processing progresses using other image understanding technologies such as shadow and sky detection, geometry recovery and intrinsic image decomposition, which limit their performances. We propose an end to end way of outdoor sun orientation estimation via a novel deep convolutional neural network (DCNN), which directly outputs the orientation of the sun from an outdoor image. Our proposed SunOriNet contains a contact layer that directly contacts the intermediate feature maps to the high-level ones and learns hierarchical features automatically from a large-scale image dataset with annotated sun orientations. The experiments reveal that our DCNN can well estimate sun orientation from a single outdoor image. The estimation accuracy of our method outperforms state-of-the-art DCNN based methods.