A system for sub-pixel disparity estimation is described herein. The system includes memory circuitry to store image data and at least one processor to execute instructions to calculate a first disparity for a set of reference views. The reference views correspond to a first subset of views among a plurality of sub-aperture views represented in the image data. The at least one processor is to refine the first disparity to a second disparity for the reference views. The second disparity has higher precision than the first disparity. The at least one processor is to map the second disparity from the reference views to a second subset of views among the plurality of sub-aperture views different than the first subset of views.

A system for sub-pixel disparity estimation is described herein. The system includes memory circuitry to store image data and at least one processor to execute instructions to calculate a first disparity for a set of reference views. The reference views correspond to a first subset of views among a plurality of sub-aperture views represented in the image data. The at least one processor is to refine the first disparity to a second disparity for the reference views. The second disparity has higher precision than the first disparity. The at least one processor is to map the second disparity from the reference views to a second subset of views among the plurality of sub-aperture views different than the first subset of views.

A method is proposed for estimating a depth for pixels in a matrix of M images. Such method comprises, at least for one set of N images among the M images, 2<N≤M, a process comprising: —determining depth maps for the images in the set of N images delivering a set of N depth maps; —for at least one current pixel for which a depth has not yet been estimated: —deciding if a candidate depth corresponding to a depth value in the set of N depth maps is consistent or not with the other depth map(s) of the set of N depth maps; —selecting the candidate depth as being the estimated depth for the current pixel if the candidate depth is decided as consistent. The process is enforced iteratively with a new N value which is lower than the previous N value used in the previous iteration of the process.

A method is proposed for estimating a depth for pixels in a matrix of M images. Such method comprises, at least for one set of N images among the M images, 2<N≤M, a process comprising: —determining depth maps for the images in the set of N images delivering a set of N depth maps; —for at least one current pixel for which a depth has not yet been estimated: —deciding if a candidate depth corresponding to a depth value in the set of N depth maps is consistent or not with the other depth map(s) of the set of N depth maps; —selecting the candidate depth as being the estimated depth for the current pixel if the candidate depth is decided as consistent. The process is enforced iteratively with a new N value which is lower than the previous N value used in the previous iteration of the process.

Disclosed are a depth refinement system and a method for a sparse depth image in a multi-aperture camera. The method includes providing a sparse depth map generated based on an image obtained through each of a plurality of apertures included in the multi-aperture camera, wherein the sparse depth map includes depths of pixels included in the image, and performing a depth noise reduction (DNR) based on the sparse depth map.

An image signal processor including a CPU is provided. The CPU receives image data and positional information of phase detection pixels from an imaging device, extracts first phase detection pixel data and second phase detection pixel data from the image data using the positional information of phase detection pixels, computes first phase graphs from the first phase detection pixel data based upon moving a first window, computes second phase graphs from the second phase detection pixel data based upon moving a second window, computes disparities of the phase detection pixels using the first phase graphs and the second phase graphs, and generates a depth map using the disparities.

An image signal processor including a CPU is provided. The CPU receives image data and positional information of phase detection pixels from an imaging device, extracts first phase detection pixel data and second phase detection pixel data from the image data using the positional information of phase detection pixels, computes first phase graphs from the first phase detection pixel data based upon moving a first window, computes second phase graphs from the second phase detection pixel data based upon moving a second window, computes disparities of the phase detection pixels using the first phase graphs and the second phase graphs, and generates a depth map using the disparities.

The disclosure provides an approach for estimating depth in a scene. According to one aspect, regions where the depth estimation is expected to perform well may first be identified in full-resolution epipolar-plane images (EPIs) generated from a plurality of images of the scene. Depth estimates for EPI-pixels with high edge confidence are determined by testing a number of discrete depth hypotheses and picking depths that lead to highest color density of sampled EPI-pixels. The depth estimate may also be propagated throughout the EPIs. This process of depth estimation and propagation may be iterated until all EPI-pixels with high edge confidence have been processed, and all EPIs may also be processed in this manner. The EPIs are then iteratively downsampled to coarser resolutions, at which edge confidence for EPI-pixels not yet processed are determined, depth estimates of EPI-pixels with high edge confidence made, and depth estimates propagated throughout the EPIs.

The disclosure provides an approach for estimating depth in a scene. According to one aspect, regions where the depth estimation is expected to perform well may first be identified in full-resolution epipolar-plane images (EPIs) generated from a plurality of images of the scene. Depth estimates for EPI-pixels with high edge confidence are determined by testing a number of discrete depth hypotheses and picking depths that lead to highest color density of sampled EPI-pixels. The depth estimate may also be propagated throughout the EPIs. This process of depth estimation and propagation may be iterated until all EPI-pixels with high edge confidence have been processed, and all EPIs may also be processed in this manner. The EPIs are then iteratively downsampled to coarser resolutions, at which edge confidence for EPI-pixels not yet processed are determined, depth estimates of EPI-pixels with high edge confidence made, and depth estimates propagated throughout the EPIs.

Methods and apparatus for estimating a depth of unfocused plenoptic data are suggested. The method includes: determining a level of homogeneity of micro-lens images of unfocused plenoptic data; determining pixels of the micro-lens images of the unfocused plenoptic unfocused plenoptic data which either have disparities equal to zero or belong to homogeneous areas as a function of the calculated level of homogeneity of the micro-lens images of the unfocused plenoptic data; and estimating the depth of the unfocused plenoptic data by a disparity estimation without considering the determined pixels. With the disclosure, by pre-processing the raw data, it can prevent any disparity estimation method to spend time on estimating disparities for: (i) pixels that are in focus, (ii) pixels that belong to homogenous areas of the scene.