A drone-projector equipped with a beam projector mounted on a drone has been investigated in order to develop a projector which can overcome restriction of place on which an image is projected. For the stability, the drone-projector requires its mass to be centered, and the additional weights related to projector should be within the payload of the drone. In addition to this requirement, the drone-projector should be designed to minimize the distortion of image caused by 3D translations or rotations of a drone during its hovering due to vibration of propellers, or global positioning system (GPS) errors. In this paper, we consider rotation of a droneprojector which makes the projected image tilted, keystoned, and shifted. To overcome this problem, we propose a software-based stabilization method which pre-corrects the image to be projected based on flight information. Our experimental results show that the distortion of the projected image due to rotations of the proposed drone-projector is attenuated by applying our stabilization method.
Auto white balance (AWB) plays a key role in a digital camera system, and determines image quality to a large extent. Existing AWB algorithms are not completely reliable and often produce inconsistent temporal estimates for consecutive frames, resulting in abrupt color changes in the output video stream. This paper presents an efficient approach to stabilize AWB estimates in a video by detecting scene and/or light source changes and adaptively setting appropriate convergence times.