Depth estimation method for light field image

Abstract

The invention discloses a depth estimation method for a light field image. The depth estimation method comprises the following steps: decoding the light field image to obtain 4D light field data and acentral sub-aperture image; generating a focus stack image sequence through the 4D light field data; drawing a focusing curve for each pixel on the central sub-aperture image; detecting a local symmetry center of the focusing curve to obtain a disparity map D1; generating an EPI image in the horizontal direction and the vertical direction corresponding to each pixel through the 4D light field data; detecting the slope of a straight line on the EPI image to obtain a disparity map D2; based on the central sub-aperture image and the two disparity maps, drawing two focusing curve segments for each pixel; calculating the matching degrees of the two focusing curve segments and the focusing curve, and determining a final disparity map D according to the matching degrees; performing deep optimization based on energy minimization. A depth estimation result of the method is high in accuracy, and has high robustness for outdoor noise scenes, complex shielding scenes and the like.

Description

technical field [0001] The invention relates to the technical fields of computer vision and digital image processing, in particular to a depth estimation method for light field images. Background technique [0002] In the first half of the 20th century, A. Gershun and others have proposed the concept of light field to describe the radiation characteristics of light in three-dimensional space. However, the light field imaging technology lags behind the development of theoretical concepts. Light field imaging devices that have emerged one after another mainly include camera arrays, light field sensors, and microlens arrays. The hand-held light field camera designed by R.Ng et al. in 2005 has become a representative of microlens array light field acquisition equipment, which directly promoted the rapid entry of light field cameras (such as Lytro, Raytrix, etc.) into the civilian market. It set off an upsurge of applied research on light field imaging. [0003] Obtaining scene...

AI Technical Summary

Problems solved by technology

[0002] In the first half of the 20th century, A. Gershun et al. have proposed the concept of light field to describe the radiation characteristics of light in three-dimensional space. However, light field imaging Technology relatively lags behind the development of theoretical concepts

Existing light-field-based depth estimation methods pay less attention to the robustness of depth estimation to noise and occlusion, which are crucial for depth estimation in complex scenes

Images