"Light-Field Dataset for Disparity Based Depth Estimation"

"A light field (LF) camera differs from a conventional camera by including a two-dimensional array of micro-lenses between the main lens and the sensor. Pixels beneath each micro-lens capture light from different sub-apertures, allowing the sensor to record both spatial and angular information of a scene. We introduce a publicly available light field (LF) image dataset designed to support the development and evaluation of disparity-based LF depth estimation algorithms. The dataset contains 285 real LF images captured using a Lytro Illum LF camera and 13 synthetic LF images. In addition, a synthetic dataset is created to replicate disparity characteristics similar to those observed in real LF cameras. The dataset also includes both real and synthetic stereo LF data, generated using a mechanical gantry system and Blender. It is intended to provide diverse scenes and controlled conditions for testing depth estimation methods and analyzing LF imaging properties."