Disparity map cavity filling method and device, electronic equipment and storage medium

Abstract

The embodiment of the invention provides a disparity map cavity filling method and device, and the method comprises the steps: determining a left effective disparity pixel of a to-be-filled pixel anda right effective disparity pixel of the to-be-filled pixel for each to-be-filled pixel; calculating a parallax difference value corresponding to a to-be-filled pixel and an original image pixel valuedifference value corresponding to the to-be-filled pixel, and calculating a parallax estimation value of the to-be-filled pixel based on the parallax difference value corresponding to the to-be-filled pixel and the original image pixel value difference value corresponding to the to-be-filled pixel; and taking the parallax estimation values of at least part of the to-be-filled pixels as parallax values of the to-be-filled pixels. Meanwhile, the relevance between the parallax value difference of the effective parallax pixels on the two sides of the to-be-filled pixel and the parallax value of the to-be-filled pixel and the relevance between the pixel value difference of the original image pixels corresponding to the effective parallax pixels on the two sides and the parallax value of the to-be-filled pixel are considered. Stripe flaws and highlight spots are avoided, and parallax image hole filling is completed in a short time.

Description

technical field [0001] The present application relates to the field of image processing, in particular to a method and device for filling holes in a disparity map, and a storage medium for electronic equipment. Background technique [0002] A disparity map is an image used to represent the translation relationship between a target image and a reference image in stereo matching technology. The steps of consistency detection, occlusion detection, and mismatch detection in stereo matching technology all rely on the disparity map. Since the reference image does not contain all the required information in the target image, resulting in holes in the target image, the disparity map includes regions to be filled corresponding to the holes in the target image. Before the step of relying on the disparity map in the stereo matching technology using the disparity map, it is necessary to fill the disparity map hole in the area to be filled in the disparity map, that is, for each pixel t...

AI Technical Summary

Problems solved by technology

The scan line filling algorithm directly selects the minimum value of the effective parallax value on the left and the effective parallax value on the right for the pixels to be filled in the area to be filled, resulting in obvious stripe defects

The variable-size average filtering algorithm is to iteratively filter the area to be filled through the average filter window from large to small to determine the disparity value of each pixel to be filled. On the one hand, for each pixel to be filled, it needs to go through The disparity value of the pixel to be filled can only be determined after multiple iterations of filtering, resulting in a long time-consuming disparity map hole filling

On the other hand, bright spots appear at pixels with wrong disparity values ​​determined

Images