RSLF+
收藏资源简介:
RSLF+数据集是由勃艮第大学、洛林大学和弗朗什-孔泰大学联合创建的,专门用于评估受滚动快门(RS)效应影响的轻场(LF)图像的密集深度估计方法。该数据集包含具有高度纹理表面和不同几何细节级别的合成RS LF图像,旨在解决在动态场景中由于RS效应导致的图像变形问题。数据集的创建过程结合了2D高斯喷射技术,通过优化相机运动参数来补偿RS效应。该数据集的应用领域主要集中在计算机视觉和图形学中,特别是在处理由RS传感器引起的图像失真问题。
The RSLF+ Dataset was co-developed by the University of Burgundy, the University of Lorraine, and the University of Franche-Comté, specifically designed to evaluate dense depth estimation methods for light field (LF) images affected by rolling shutter (RS) effects. This dataset contains synthetic RS LF images with highly textured surfaces and varying levels of geometric details, aiming to resolve image distortion issues caused by RS effects in dynamic scenes. The dataset's development integrates 2D Gaussian splatting technology, with camera motion parameters optimized to compensate for RS effects. The application scenarios of this dataset are primarily focused on computer vision and graphics, especially in handling image distortion problems induced by RS sensors.

- 1Dense Scene Reconstruction from Light-Field Images Affected by Rolling Shutter勃艮第大学, 洛林大学, 弗朗什-孔泰大学 · 2024年



