High-frequency shape recovery from shading by CNN and domain adaptation
This work addresses a limitation in one-shot scanning for moving objects, offering an incremental improvement in shape recovery.
The paper tackles the problem of recovering high-frequency shapes from sparse structured-light scans by using shading information from a single RGB-D image, proposing a learning-based approach with domain adaptation for data augmentation, and demonstrates its effectiveness experimentally.
Importance of structured-light based one-shot scanning technique is increasing because of its simple system configuration and ability of capturing moving objects. One severe limitation of the technique is that it can capture only sparse shape, but not high frequency shapes, because certain area of projection pattern is required to encode spatial information. In this paper, we propose a technique to recover high-frequency shapes by using shading information, which is captured by one-shot RGB-D sensor based on structured light with single camera. Since color image comprises shading information of object surface, high-frequency shapes can be recovered by shape from shading techniques. Although multiple images with different lighting positions are required for shape from shading techniques, we propose a learning based approach to recover shape from a single image. In addition, to overcome the problem of preparing sufficient amount of data for training, we propose a new data augmentation method for high-frequency shapes using synthetic data and domain adaptation. Experimental results are shown to confirm the effectiveness of the proposed method.